1. A server with maintenance prediction functionality, comprising:

the acquisition module is used for acquiring parameter data for predictive maintenance of the wave power generator;

the fault characteristic frequency acquisition module is used for acquiring the fault characteristic frequency of the wave generator bearing according to the parameter data;

the power spectral density acquisition module is used for acquiring a cyclic autocorrelation function and power spectral density of the vibration signal in the parameter data according to the fault characteristic frequency;

the optimal projection acquisition module is used for acquiring an optimal projection direction of a one-dimensional space according to sample data of power spectral density;

the operation health degree obtaining module is used for predicting the operation health degree of the wave power generator according to the optimal projection direction of the one-dimensional space; and

and the early warning module is used for early warning the running of the wave-activated generator according to the running health degree of the wave-activated generator.

2. The server with maintenance prediction function according to claim 1,

the acquisition module is adapted to acquire parameter data for predictive maintenance of the wave-activated generator, i.e.

And acquiring a vibration signal, the rpm of the working shaft, the diameter of the rolling body, the diameter of the bearing joint, the pressure angle of the bearing and the number of moving bodies of the wave-activated generator.

3. The server with maintenance prediction function according to claim 2,

the fault characteristic frequency acquisition module is suitable for acquiring the fault characteristic frequency of the bearing of the wave power generator according to the parameter data, namely

Acquiring the characteristic frequency of the fault of the inner ring of the bearing of the wave power generator:

acquiring the outer ring fault characteristic frequency of a bearing of the wave power generator:

acquiring the fault characteristic frequency of a rolling body of a bearing of the wave power generator;

wherein N is the rotating speed of the working shaft per minute; d is the diameter of the rolling body; d is the diameter of the bearing section; alpha is a bearing pressure angle; z is the number of rolling elements.

4. The server with maintenance prediction function according to claim 3,

the power spectral density acquisition module is adapted to acquire the power spectral density and the cyclic autocorrelation function of the vibration signal in the parametric data based on the characteristic frequency of the fault, i.e. the power spectral density

Acquiring a cyclic autocorrelation function and a power spectrum of the vibration signal when the cyclic frequency is equal to the fault characteristic frequency;

performing wavelet decomposition on the vibration signal with the cycle frequency equal to the fault characteristic frequency according to a preset scale;

carrying out threshold value denoising processing on the high-frequency coefficient subjected to wavelet decomposition;

reconstructing signals by using the low-frequency and high-frequency coefficients with the same number of layers as the preset scale after wavelet decomposition and denoising;

selecting a circulation frequency equal to the fault characteristic frequency and calculating a circulation autocorrelation function value of a reconstructed signal at the circulation frequency;

performing modular operation on the cyclic autocorrelation function value;

and acquiring the power spectral density of the modulus value of the cyclic autocorrelation function.

5. The server with maintenance prediction function according to claim 4,

the optimal projection acquisition module is adapted to acquire an optimal projection direction of a one-dimensional space, i.e. an optimal projection direction of a power spectral density, from sample data of the sample data

Establishing a data vector:

x＝(x⁽¹⁾,x⁽²⁾,x⁽³⁾)；

wherein x is⁽¹⁾The power spectrum density of the amplitude of the cyclic autocorrelation function at the cyclic frequency equal to the characteristic frequency of the inner ring fault is taken as the cyclic frequency; x is the number of⁽²⁾The power spectrum density of the amplitude of the cyclic autocorrelation function at the cyclic frequency equal to the characteristic frequency of the outer ring fault is taken as the cyclic frequency; x is the number of⁽³⁾The power spectral density of the amplitude of the cyclic autocorrelation function at the cyclic frequency equal to the fault characteristic frequency of the rolling body is taken as the reference power spectral density;

obtain transform vector w ═ w⁽¹⁾,w⁽²⁾,w⁽³⁾)；

Wherein, w⁽¹⁾The power spectrum density coefficient of the amplitude of the cyclic autocorrelation function at the cyclic frequency equal to the characteristic frequency of the inner ring fault; w is a⁽²⁾The power spectrum density coefficient of the amplitude of the cyclic autocorrelation function at the cyclic frequency equal to the outer ring fault characteristic frequency is taken as the cyclic frequency; w is a⁽³⁾The power spectral density coefficient of the amplitude of the cyclic autocorrelation function at the cyclic frequency equal to the fault characteristic frequency of the rolling body is taken as the reference value;

obtaining an optimal transformation vector w^*I.e. by

Then, w when J (w) takes the maximum value is the optimal transformation vector, i.e. w^*；

S_B＝(m₂-m₁)(m₂-m₁)^T；

Wherein, C₁A set of data samples corresponding to the type of the wave power generator working normally; c₂A set of data samples corresponding to the type of the wave power generator with faults; x is the number of_aIs C₁A middle a data sample; x is the number of_bIs C₁The (b) th data sample; n is a radical of₁Is C₁Total number of data samples for a category; n is a radical of₂Is C₂Total number of data samples for a category; s₁Is C₁Intra-class variance of the class; s₂Is C₂Intra-class variance of the class; m is₁Is C₁A mean vector of categories; m is₂Is C₂A mean vector of categories; s_BIs an inter-class covariance matrix; s_wIs an intra-class covariance matrix; t is transposition;

optimal transformation vector w^*Is the projection line direction, i.e. the optimal projection direction of the sample data to the one-dimensional space.

6. The server with maintenance prediction function according to claim 5,

the operation health degree acquisition module is suitable for predicting the operation health degree of the wave-activated generator according to the optimal projection direction of the one-dimensional space, namely

Acquiring mean values and variances of data of two categories, namely, failed data and normal data, which are obtained after the sample data is projected in the optimal projection direction;

μ₁＝w^*Tm₁；

μ₂＝w^*Tm₂；

wherein, mu₁Data in the normal category of operation is at w^*Mean value of the data obtained after vector axis projection; mu.s₂For the fault category data at w^*Mean value of the data obtained after vector axis projection; delta₁Data in the normal category of operation is at w^*Standard deviation of data obtained after vector axis projection; delta₂For the fault category data at w^*Standard deviation of data obtained after vector axis projection;

then the corresponding variance can be obtained according to the standard deviation;

predicting the health of the operation of the wave-power generator, i.e.

The current data is x_c；

Z_c＝w^*x_c；

Wherein Z is_cFor current data as x_cAt w^*The value obtained after vector axis projection;

the current running health J of the wave power generator is:

when Z is_c≤μ₁+δ₁When the wave power generator operates, the operation health degree J is 0;

when Z is_c≥μ₂-δ₂Meanwhile, the running health degree J of the wave power generator is 1;

when mu is₁+δ₁≤Z_c≤μ₂-δ₂In the meantime, the running health degree J of the wave power generator is as follows:

the running health degree J of the wave-activated generator is a number between 0 and 1, and the working state of the wave-activated generator is healthier when the numerical value is larger and is closer to 1; smaller values and closer values to 0 indicate a less healthy operating condition of the wave generator.

7. The server with maintenance prediction function according to claim 6,

the early warning module is suitable for early warning the running of the wave-activated generator according to the running health degree of the wave-activated generator, namely

When J is<J₀Sending out early warning information for prompting that the wave power generator needs to be maintained;

wherein, J₀Is a set early warning threshold value;

when the wave power generator cannot be maintained after maintenance early warning information is sent out, the wave power generator enters a low working load state, and the time for the floating body to be folded and put down is controlled to change the working time of the floating body into that:

Time＝J*T₀；

wherein, T₀The working time of the floating body is set when the wave power generator works normally; time is the float operating Time set under the current wave power generator operating health conditions.

8. A predictive wave generator maintenance system, comprising:

a wave power generator and a cloud server;

the wave power generator is suitable for acquiring parameter data of the body and sending the parameter data to the cloud server;

the cloud server is suitable for predicting the running health degree of the wave-activated generator according to the parameter data and carrying out early warning according to the running health degree of the wave-activated generator.

Background

The wave power generator is one of key parts of a wave power generation system, has high fault shutdown probability and is very important for early fault diagnosis of the wave power generator, but because the wave power generator generally depends on the seaside or on the sea, the environment is complex and changeable, the required performance is high, the maintenance cost is higher when the wave power generator is far away from the land, and the fault diagnosis of the wave power generator has the characteristics of small amplitude, instability, easiness in being influenced by working conditions and the like. The effective early fault online diagnosis method for the wave-activated generator is sought, the wave-activated generator is prevented from being stopped and maintained due to faults, the operation cost is increased, and the method has great significance.

The bearing is a key part of the wave power generator and plays a role of supporting a rotating structure of a mechanical system, and the failure of the bearing easily causes the abrasion of other important parts in the mechanical system. Failures of motor rolling bearings typically include wear and damage failures. In the event of a wear failure, the clearances of the motor bearing components increase as the wear increases, which may lead to increased vibration of the bearing during operation. At this moment, the vibration signal of the motor bearing can be randomly changed, and no uniform change rule can be followed. The damage failure is a failure caused by damage such as metal peeling, pitting, or galling on the motor bearing surface. The basic characteristics of the motor bearing fault are as follows: rolling of the element across the surface producing the damage produces an abrupt, shock-like pulse signal, resulting in resonance of the motor bearings and other elements.

The regular maintenance mode is carried out in a static state that the equipment does not operate, and has the following defects: the difference between the state of the equipment and the operation under the condition of no work is obvious, and the judgment accuracy is influenced. Because the equipment is regularly checked and maintained, even if the equipment is in a good state, the equipment still needs to be tested and maintained according to a plan, and manpower and material resources are wasted.

Therefore, it is necessary to design a new server and system with a maintenance prediction function based on the above technical problems.

Disclosure of Invention

The invention aims to provide a server with a maintenance prediction function and a system.

In order to solve the above technical problem, the present invention provides a server with a maintenance prediction function, including:

the acquisition module is used for acquiring parameter data for predictive maintenance of the wave power generator;

the fault characteristic frequency acquisition module is used for acquiring the fault characteristic frequency of the wave generator bearing according to the parameter data;

the power spectral density acquisition module is used for acquiring a cyclic autocorrelation function and power spectral density of the vibration signal in the parameter data according to the fault characteristic frequency;

the optimal projection acquisition module is used for acquiring an optimal projection direction of a one-dimensional space according to sample data of power spectral density;

the operation health degree obtaining module is used for predicting the operation health degree of the wave power generator according to the optimal projection direction of the one-dimensional space; and

and the early warning module is used for early warning the running of the wave-activated generator according to the running health degree of the wave-activated generator.

Further, the acquisition module is adapted to acquire parameter data for predictive maintenance of the wave power generator, i.e.

And acquiring a vibration signal, the rpm of the working shaft, the diameter of the rolling body, the diameter of the bearing joint, the pressure angle of the bearing and the number of moving bodies of the wave-activated generator.

Further, the fault signature frequency acquisition module is adapted to acquire a fault signature frequency of the wave generator bearing according to the parameter data, i.e. the fault signature frequency is acquired by the fault signature frequency acquisition module

Acquiring the characteristic frequency of the fault of the inner ring of the bearing of the wave power generator:

acquiring the outer ring fault characteristic frequency of a bearing of the wave power generator:

acquiring the fault characteristic frequency of a rolling body of a bearing of the wave power generator;

wherein N is the rotating speed of the working shaft per minute; d is the diameter of the rolling body; d is the diameter of the bearing section; alpha is a bearing pressure angle; z is the number of rolling elements.

Further, the power spectral density acquisition module is adapted to acquire the power spectral density and the cyclic autocorrelation function of the vibration signal in the parametric data based on the characteristic frequency of the fault, i.e. the power spectral density

Acquiring a cyclic autocorrelation function and a power spectrum of the vibration signal when the cyclic frequency is equal to the fault characteristic frequency;

performing wavelet decomposition on the vibration signal with the cycle frequency equal to the fault characteristic frequency according to a preset scale;

carrying out threshold value denoising processing on the high-frequency coefficient subjected to wavelet decomposition;

reconstructing signals by using the low-frequency and high-frequency coefficients with the same number of layers as the preset scale after wavelet decomposition and denoising;

selecting a circulation frequency equal to the fault characteristic frequency and calculating a circulation autocorrelation function value of a reconstructed signal at the circulation frequency;

performing modular operation on the cyclic autocorrelation function value;

and acquiring the power spectral density of the modulus value of the cyclic autocorrelation function.

Further, the optimal projection acquisition module is adapted to acquire an optimal projection direction of a one-dimensional space, i.e. an optimal projection direction of a power spectral density, from sample data of the sample data

Establishing a data vector:

x＝(x⁽¹⁾,x⁽²⁾,x⁽³⁾)；

wherein x is⁽¹⁾Is a cyclic frequencyThe frequency is equal to the power spectral density of the amplitude of the cyclic autocorrelation function at the inner ring fault characteristic frequency; x is the number of⁽²⁾The power spectrum density of the amplitude of the cyclic autocorrelation function at the cyclic frequency equal to the characteristic frequency of the outer ring fault is taken as the cyclic frequency; x is the number of⁽³⁾The power spectral density of the amplitude of the cyclic autocorrelation function at the cyclic frequency equal to the fault characteristic frequency of the rolling body is taken as the reference power spectral density;

obtain transform vector w ═ w⁽¹⁾,w⁽²⁾,w⁽³⁾)；

Wherein, w⁽¹⁾The power spectrum density coefficient of the amplitude of the cyclic autocorrelation function at the cyclic frequency equal to the characteristic frequency of the inner ring fault; w is a⁽²⁾The power spectrum density coefficient of the amplitude of the cyclic autocorrelation function at the cyclic frequency equal to the outer ring fault characteristic frequency is taken as the cyclic frequency; w is a⁽³⁾The power spectral density coefficient of the amplitude of the cyclic autocorrelation function at the cyclic frequency equal to the fault characteristic frequency of the rolling body is taken as the reference value;

obtaining an optimal transformation vector w^*I.e. by

Then, w when J (w) takes the maximum value is the optimal transformation vector, i.e. w^*；

S_B＝(m₂-m₁)(m₂-m₁)^T；

Wherein, C₁A set of data samples corresponding to the type of the wave power generator working normally; c₂As wave-activated generatorsThe type of the fault corresponds to a set of data samples; x is the number of_aIs C₁A middle a data sample; x is the number of_bIs C₁The (b) th data sample; n is a radical of₁Is C₁Total number of data samples for a category; n is a radical of₂Is C₂Total number of data samples for a category; s₁Is C₁Intra-class variance of the class; s₂Is C₂Intra-class variance of the class; m is₁Is C₁A mean vector of categories; m is₂Is C₂A mean vector of categories; s_BIs an inter-class covariance matrix; s_wIs an intra-class covariance matrix; t is transposition;

optimal transformation vector w^*Is the projection line direction, i.e. the optimal projection direction of the sample data to the one-dimensional space.

Further, the operation health degree acquisition module is adapted to predict the operation health degree of the wave-activated generator according to the optimal projection direction of the one-dimensional space, i.e. the operation health degree of the wave-activated generator is predicted according to the optimal projection direction of the one-dimensional space

Acquiring mean values and variances of data of two categories, namely, failed data and normal data, which are obtained after the sample data is projected in the optimal projection direction;

μ₁＝w^*Tm₁；

μ₂＝w^*Tm₂；

wherein, mu₁Data in the normal category of operation is at w^*Mean value of the data obtained after vector axis projection; mu.s₂For the fault category data at w^*Obtained after projection of vector axisThe mean of the data of (a); delta₁Data in the normal category of operation is at w^*Standard deviation of data obtained after vector axis projection; delta₂For the fault category data at w^*Standard deviation of data obtained after vector axis projection;

then the corresponding variance can be obtained according to the standard deviation;

predicting the health of the operation of the wave-power generator, i.e.

The current data is x_c；

Z_c＝w^*x_c；

Wherein Z is_cIs the preceding data x_cAt w^*The value obtained after vector axis projection;

the current running health J of the wave power generator is:

when Z is_c≤μ₁+δ₁When the wave power generator operates, the operation health degree J is 0;

when Z is_c≥μ₂-δ₂Meanwhile, the running health degree J of the wave power generator is 1;

when mu is₁+δ₁≤Z_c≤μ₂-δ₂In the meantime, the running health degree J of the wave power generator is as follows:

the running health degree J of the wave-activated generator is a number between 0 and 1, and the working state of the wave-activated generator is healthier when the numerical value is larger and is closer to 1; smaller values and closer values to 0 indicate a less healthy operating condition of the wave generator.

Further, the early warning module is suitable for carrying out early warning on the running of the wave-activated generator according to the running health degree of the wave-activated generator, namely

When J is<J₀Sending out early warning information for prompting that the wave power generator needs to be maintained;

wherein, J₀Is a set early warning threshold value;

when the wave power generator cannot be maintained after maintenance early warning information is sent out, the wave power generator enters a low working load state, and the time for the floating body to be folded and put down is controlled to change the working time of the floating body into that:

Time＝J*T₀；

wherein, T₀The working time of the floating body is set when the wave power generator works normally; time is the float operating Time set under the current wave power generator operating health conditions.

In a second aspect, the present invention also provides a wave generator predictive maintenance system comprising:

a wave power generator and a cloud server;

the wave power generator is suitable for acquiring parameter data of the body and sending the parameter data to the cloud server;

the cloud server is suitable for predicting the running health degree of the wave-activated generator according to the parameter data and carrying out early warning according to the running health degree of the wave-activated generator.

The invention has the advantages that parameter data for predictive maintenance of the wave-activated generator are collected through the collection module; the fault characteristic frequency acquisition module is used for acquiring the fault characteristic frequency of the wave generator bearing according to the parameter data; the power spectral density acquisition module is used for acquiring a cyclic autocorrelation function and power spectral density of the vibration signal in the parameter data according to the fault characteristic frequency; the optimal projection acquisition module is used for acquiring an optimal projection direction of a one-dimensional space according to sample data of power spectral density; the operation health degree obtaining module is used for predicting the operation health degree of the wave power generator according to the optimal projection direction of the one-dimensional space; and the early warning module is used for early warning the running of the wave-activated generator according to the running health degree of the wave-activated generator, realizing the prediction and early warning of the fault of the wave-activated generator, and overcoming the defects of strong subjectivity of manual maintenance and high labor cost.

Additional features and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention will be realized and attained by the structure particularly pointed out in the written description and drawings.

In order to make the aforementioned and other objects, features and advantages of the present invention comprehensible, preferred embodiments accompanied with figures are described in detail below.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.

FIG. 1 is a functional block diagram of a server with maintenance prediction functionality in accordance with the present invention;

fig. 2 is a schematic block diagram of a predictive maintenance system for a wave power generator according to the present invention.

Detailed Description

To make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings, and it is apparent that the described embodiments are some, but not all embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

Example 1

Fig. 1 is a schematic block diagram of a server with a maintenance prediction function according to the present invention.

As shown in fig. 1, this embodiment 1 provides a server with a maintenance prediction function, including: the acquisition module is used for acquiring parameter data for predictive maintenance of the wave power generator; the fault characteristic frequency acquisition module is used for acquiring the fault characteristic frequency of the wave generator bearing according to the parameter data; the power spectral density acquisition module is used for acquiring a cyclic autocorrelation function and power spectral density of the vibration signal in the parameter data according to the fault characteristic frequency; the optimal projection acquisition module is used for acquiring an optimal projection direction of a one-dimensional space according to sample data of power spectral density; the operation health degree obtaining module is used for predicting the operation health degree of the wave power generator according to the optimal projection direction of the one-dimensional space; the early warning module is used for early warning the running of the wave power generator according to the running health degree of the wave power generator, so that the fault of the wave power generator is predicted and early warned, and the defects of strong subjectivity of manual maintenance and high labor cost are overcome; the online monitoring is carried out on the wave-activated generator, the running health degree information of the wave-activated generator can be given in real time according to the current parameter data, the fastest and most accurate predictive maintenance information is provided for the online predictive maintenance of the wave-activated generator, the predictive maintenance is carried out once the early warning condition is met, the early warning is started and maintained, and the wave-activated generator is started to enter a low working load state, so that the increase of the operation cost caused by the shutdown maintenance of the sudden fault of the wave-activated generator is avoided, and the defects of strong subjectivity in using manpower for maintenance and high labor cost are overcome.

In this embodiment, the acquisition module is adapted to acquire parameter data for predictive maintenance of the wave-activated generator, that is, acquiring vibration signals, the rpm of the working shaft, the diameter of the rolling element, the diameter of the bearing node, the pressure angle of the bearing, and the number of moving bodies of the wave-activated generator; the vibration signal of the wave power generator can be acquired by an acceleration sensor arranged on the wave power generator.

In this embodiment, the fault characteristic frequency obtaining module is adapted to obtain the fault characteristic frequency of the wave generator bearing according to the parameter data, that is, obtain the inner ring fault characteristic frequency of the wave generator bearing:

acquiring the outer ring fault characteristic frequency of a bearing of the wave power generator:

acquiring the fault characteristic frequency of a rolling body of a bearing of the wave power generator;

wherein N is the rotating speed of the working shaft per minute; d is the diameter of the rolling body; d is the diameter of the bearing section; alpha is a bearing pressure angle; z is the number of rolling elements.

In this embodiment, the power spectral density obtaining module is adapted to obtain the cyclic autocorrelation function and the power spectral density of the vibration signal in the parameter data according to the fault characteristic frequency, that is, the cyclic autocorrelation function and the power spectral density of the vibration signal when the cyclic frequency is equal to the fault characteristic frequency may be calculated by the cloud server; acquiring a cyclic autocorrelation function and a power spectrum of the vibration signal when the cyclic frequency is equal to the fault characteristic frequency; performing wavelet decomposition on a vibration signal with a cycle frequency equal to a fault characteristic frequency according to a preset scale, taking the scale s as 4 in order to reduce the influence of noise and avoid losing information, namely performing 4-layer decomposition on the signal; performing threshold value denoising processing on the high-frequency coefficient subjected to wavelet decomposition, wherein a noise signal is represented as a high-frequency signal, and denoising mainly aims at the high-frequency coefficient; reconstructing signals by using the low-frequency and high-frequency coefficients with the same number of layers as the preset scale after wavelet decomposition and denoising, namely reconstructing signals by using the decomposed and denoised 4 th layer low-frequency and high-frequency coefficients; selecting a circulation frequency equal to the fault characteristic frequency and calculating a circulation autocorrelation function value of a reconstructed signal at the circulation frequency; performing modular operation on the cyclic autocorrelation function value; and acquiring the power spectral density of the modulus value of the cyclic autocorrelation function.

In this embodiment, the optimal projection obtaining module is adapted to obtain an optimal projection direction of a one-dimensional space according to sample data of power spectral density, that is, an optimal projection direction of a sample data feature projected to a one-dimensional space, that is, a Fisher method:

establishing a data vector:

x＝(x⁽¹⁾,x⁽²⁾,x⁽³⁾)；

wherein x is⁽¹⁾For the amplitude of the cyclic autocorrelation function at a cyclic frequency equal to the characteristic frequency of the inner ring faultA value power spectral density; x is the number of⁽²⁾The power spectrum density of the amplitude of the cyclic autocorrelation function at the cyclic frequency equal to the characteristic frequency of the outer ring fault is taken as the cyclic frequency; x is the number of⁽³⁾The power spectral density of the amplitude of the cyclic autocorrelation function at the cyclic frequency equal to the fault characteristic frequency of the rolling body is taken as the reference power spectral density;

obtain transform vector w ═ w⁽¹⁾,w⁽²⁾,w⁽³⁾)；

Wherein, w⁽¹⁾The power spectrum density coefficient of the amplitude of the cyclic autocorrelation function at the cyclic frequency equal to the characteristic frequency of the inner ring fault; w is a⁽²⁾The power spectrum density coefficient of the amplitude of the cyclic autocorrelation function at the cyclic frequency equal to the outer ring fault characteristic frequency is taken as the cyclic frequency; w is a⁽³⁾The power spectral density coefficient of the amplitude of the cyclic autocorrelation function at the cyclic frequency equal to the fault characteristic frequency of the rolling body is taken as the reference value; each cyclic autocorrelation function amplitude power spectrum density coefficient is obtained according to the calendar data of the corresponding cyclic autocorrelation function amplitude power spectrum density and is a real number;

obtaining an optimal transformation vector w^*After sample data is projected, various samples are separated as far as possible in a one-dimensional space, the larger the difference between the two mean values is, the better the difference is, the denser the inside of various samples is, the smaller the dispersion in the samples is, the better the dispersion in the samples is, namely, the samples are obtained

Then, w when J (w) takes the maximum value is the optimal transformation vector, i.e. w^*；

S_B＝(m₂-m₁)(m₂-m₁)^T；

Wherein, C₁A set of data samples corresponding to the type of the wave power generator working normally; c₂A set of data samples corresponding to the type of the wave power generator with faults; x is the number of_aIs C₁A middle a data sample; x is the number of_bIs C₁The (b) th data sample; n is a radical of₁Is C₁Total number of data samples for a category; n is a radical of₂Is C₂Total number of data samples for a category; s₁Is C₁Intra-class variance of the class; s₂Is C₂Intra-class variance of the class; m is₁Is C₁A mean vector of categories; m is₂Is C₂A mean vector of categories; s_BIs an inter-class covariance matrix; s_wIs an intra-class covariance matrix; t is transposition; the data sample is a data vector constructed by the data of the corresponding type wave power generator;

solving by using a Lagrange multiplier method to obtain:

optimal transformation vector w^*Is the projection line direction, i.e. the optimal projection direction of the sample data to the one-dimensional space.

In this embodiment, the operation health degree obtaining module is adapted to predict the operation health degree of the wave-activated generator according to the optimal projection direction of the one-dimensional space, that is, obtain the mean and variance of data of two categories, namely, data obtained after projection of sample data in the optimal projection direction, which has a fault and works normally;

μ₁＝w^*Tm₁；

μ₂＝w^*Tm₂；

wherein, mu₁Data in the normal category of operation is at w^*Mean value of the data obtained after vector axis projection; mu.s₂For the fault category data at w^*Mean value of the data obtained after vector axis projection; delta₁Data in the normal category of operation is at w^*Standard deviation of data obtained after vector axis projection; delta₂For the fault category data at w^*Standard deviation of data obtained after vector axis projection;

then the corresponding variance can be obtained according to the standard deviation;

predicting the health of the operation of the wave-power generator, i.e.

Let current data be x_c；

Z_c＝w^*x_c；

Wherein Z is_cIs the preceding data x_cAt w^*The value obtained after vector axis projection; the current data is a data vector constructed by real-time data of the wave power generator;

the current running health J of the wave power generator is:

when Z is_c≤μ₁+δ₁When the wave power generator operates, the operation health degree J is 0;

when Z is_c≥μ₂-δ₂Meanwhile, the running health degree J of the wave power generator is 1;

when mu is₁+δ₁≤Z_c≤μ₂-δ₂In the meantime, the running health degree J of the wave power generator is as follows:

the running health degree J of the wave-activated generator is a number between 0 and 1, and the working state of the wave-activated generator is healthier when the numerical value is larger and is closer to 1; smaller values and closer values to 0 indicate a less healthy operating condition of the wave generator.

In this embodiment, the early warning module is suitable for early warning the running of the wave power generator according to the running health degree of the wave power generator, namely

When J is<J₀Sending out early warning information for prompting that the wave power generator needs to be maintained;

wherein, J₀For a set warning threshold, for example, set to 0.6;

when the wave power generator cannot be maintained after maintenance early warning information is sent out, the wave power generator enters a low working load state, and the time for the floating body to be folded and put down is controlled to change the working time of the floating body into that:

Time＝J*T₀；

wherein, T₀The working time of the floating body is set when the wave power generator works normally; time is the float operating Time set under the current wave power generator operating health conditions.

Example 2

Fig. 2 is a schematic block diagram of a predictive maintenance system for a wave power generator according to the present invention.

As shown in fig. 2, based on embodiment 1, this embodiment 2 further provides a wave power generator predictive maintenance system, including: a wave power generator and a cloud server; the wave power generator is suitable for acquiring parameter data of the body and sending the parameter data to the cloud server; the cloud server is suitable for predicting the running health degree of the wave-activated generator according to the parameter data and carrying out early warning according to the running health degree of the wave-activated generator.

In this embodiment, the cloud server is adapted to use the server with the maintenance prediction function in embodiment 1 to predict the operation health degree of the wave power generator, and perform early warning on the operation of the wave power generator according to the operation health degree of the wave power generator.

In this embodiment, the wave power generator comprises: the device comprises a control module, a detection module and a communication module, wherein the detection module and the communication module are electrically connected with the control module; the detection module is suitable for detecting vibration signals of the wave power generator, and the control module is suitable for sending the vibration signals of the wave power generator to the cloud server through the communication module; the detection module comprises: a vibration sensor, a signal amplification circuit and an AD converter; the vibration signal detected by the vibration sensor is amplified by the signal amplifying circuit, converted into a digital signal by the AD converter and then input into the control module; the control module can control a floating body in the wave power generator; the control module can be but is not limited to an ARM microprocessor; the vibration sensor may be, but is not limited to, an acceleration sensor; the server can obtain a wave power generator running state monitoring model after training and testing according to parameter data, namely a mathematical model adopted by the server with a maintenance prediction function in embodiment 1, predict the possible degree of the current wave power generator failure by combining with real-time vibration signals, set a threshold value of maintenance early warning, and start the maintenance early warning or control the effective working time of the floating body once the early warning condition is met, so that the wave power generator enters a low working load state.

In summary, the invention collects the parameter data for predictive maintenance of the wave power generator through the collection module; the fault characteristic frequency acquisition module is used for acquiring the fault characteristic frequency of the wave generator bearing according to the parameter data; the power spectral density acquisition module is used for acquiring a cyclic autocorrelation function and power spectral density of the vibration signal in the parameter data according to the fault characteristic frequency; the optimal projection acquisition module is used for acquiring an optimal projection direction of a one-dimensional space according to sample data of power spectral density; the operation health degree obtaining module is used for predicting the operation health degree of the wave power generator according to the optimal projection direction of the one-dimensional space; and the early warning module is used for early warning the running of the wave-activated generator according to the running health degree of the wave-activated generator, realizing the prediction and early warning of the fault of the wave-activated generator, and overcoming the defects of strong subjectivity of manual maintenance and high labor cost.

In the embodiments provided in the present application, it should be understood that the disclosed apparatus and method can be implemented in other ways. The apparatus embodiments described above are merely illustrative, and for example, the flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of apparatus, methods and computer program products according to various embodiments of the present invention. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.

In addition, the functional modules in the embodiments of the present invention may be integrated together to form an independent part, or each module may exist separately, or two or more modules may be integrated to form an independent part.

The functions, if implemented in the form of software functional modules and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a cloud server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.

In light of the foregoing description of the preferred embodiment of the present invention, many modifications and variations will be apparent to those skilled in the art without departing from the spirit and scope of the invention. The technical scope of the present invention is not limited to the content of the specification, and must be determined according to the scope of the claims.