1. An interactive multi-model algorithm based on support vector regression is characterized by comprising the following steps:

s1, constructing an interactive multi-model algorithm based on support vector regression, connecting the support vector regression network serving as a feedback network with the traditional interactive multi-model algorithm, and obtaining an adjusting coefficient of the algorithm;

s2, selecting a training set and a label set, and preprocessing the training set and the label set according to a set normalization standard rule;

s3, training the support vector regression network model by using the training parameters according to the support vector regression network model and the training parameters thereof to finally achieve convergence;

and S4, inputting the target state input interaction value and the measurement of the corresponding model to be processed into the improved model to obtain the target current time state estimation value.

2. The support vector regression-based interactive multi-model algorithm of claim 1, wherein: in step S1, at each tracking time, the target state input interaction value and the measurement corresponding to each model first enter the feedback network to obtain a coefficient for adjusting the process noise covariance of different models at the current time, and the coefficient is used as an adjustment coefficient.

3. The support vector regression-based interactive multi-model algorithm of claim 1, wherein: in step S1, the number of input nodes of the vector regression network corresponds to the sum of the target state and the measurement dimension, the number of output layer nodes is 1, and the network relationship is:

4. the support vector regression-based interactive multi-model algorithm of claim 1, wherein: the training set comprises a target data set with a plurality of maneuvering grades, and the target data set is segmented to obtain the training set and a label set, wherein the label set is the target real state at each moment.

5. The support vector regression-based interactive multi-model algorithm of claim 4, wherein: the normalization standardization rule is a Min-Max standardization mode.

6. The support vector regression-based interactive multi-model algorithm of claim 1, wherein: in step S3, the MSE is used as a loss function and the back propagation neural network is trained in an end-to-end manner, where the loss function is:

where m is the number of training data.

7. An interactive multi-model system based on support vector regression, characterized by: the system comprises:

the model construction module is used for constructing an interactive multi-model algorithm based on support vector regression, and connecting a support vector regression network serving as a feedback network with a traditional interactive multi-model algorithm to obtain an adjusting coefficient of the algorithm;

the data preprocessing module is used for selecting a training set and a label set and preprocessing the training set and the label set according to a set normalization standard rule;

the network training module is used for training the support vector regression network model by using the training parameters according to the support vector regression network model and the training parameters thereof so as to finally achieve convergence; and;

and the estimation value calculation module is used for inputting the target state input interaction value and the measurement of the corresponding model to be processed into the improved model to obtain the target current time state estimation value.

8. An interactive multi-model storage medium based on support vector regression, characterized by: the storage medium having stored thereon a computer program enabling, when executed by a processor, the steps of an interactive multi-model algorithm using support vector regression according to any one of claims 1 to 7.

9. An interactive multi-model computer device based on support vector regression, characterized by: the apparatus comprises a memory, a processor, and an algorithm program using an interactive multi-model algorithm with support vector regression stored on the memory and executable on the processor, the algorithm program using an interactive multi-model algorithm with support vector regression being configured to implement the steps of the interactive multi-model algorithm with support vector regression of any one of claims 1-7.

Background

In recent years, with the development of technologies, the related performance of the unmanned aerial vehicle is greatly improved, various highly mobile unmanned aerial vehicles are continuously developed, which puts higher requirements on a target tracking technology, strong mobile refers to the relative instantaneous and violent change of target acceleration or motion mode and the continuous change of speed, angle and acceleration, and the traditional unmanned aerial vehicle tracking algorithm model has poor adaptability and large calculation amount.

The publication number CN109491241A provides a robust tracking method for Unmanned Aerial Vehicle (UAV) for maneuvering targets, which obtains a final target state estimation under an interactive multi-model framework, and the performance of the used interactive multi-model algorithm depends on the used model to a large extent, so in order to improve the filtering accuracy of the algorithm, it is necessary to cover as many motion models as possible, but this brings about a problem that the calculation amount of the algorithm is doubled. In addition, too many models in the model set may result in competition between models, thereby reducing the accuracy of the algorithm. Therefore, a model set with a proper scale is generally established in advance, and after the model set is determined, the model set is not changed in the tracking process. However, as the control technology is developed, the mobility of various targets is better and better, and it is difficult for a preset fixed number of model sets to meet the actual demand.

Disclosure of Invention

The present invention provides an interactive multi-model algorithm, system, storage medium and computer device based on support vector regression, so as to solve the problems proposed in the background art.

In order to achieve the purpose, the invention provides the following technical scheme:

an interactive multi-model algorithm based on support vector regression, comprising a bed-ridden subject, a floor and a bed, wherein:

s1, constructing an interactive multi-model algorithm based on support vector regression, connecting the support vector regression network serving as a feedback network with the traditional interactive multi-model algorithm, and obtaining an adjusting coefficient of the algorithm;

s2, selecting a training set and a label set, and preprocessing the training set and the label set according to a set normalization standard rule;

s3, training the support vector regression network model by using the training parameters according to the support vector regression network model and the training parameters thereof to finally achieve convergence;

and S4, inputting the target state input interaction value and the measurement of the corresponding model to be processed into the improved model to obtain the target current time state estimation value.

Preferably, in step S1, at each tracking time, the target state input interaction value and the measurement corresponding to each model first enter the feedback network to obtain a coefficient for adjusting the process noise covariance of different models at the current time, and the coefficient is used as an adjustment coefficient.

Preferably, in step S1, the number of input nodes of the vector regression network corresponds to the sum of the target state and the measurement dimension, the output layer node is 1, and the network relationship is:

preferably, the training set includes a target data set of multiple maneuver levels, and the target data set is segmented to obtain the training set and a label set, where the label set is a target real state at each time.

Preferably, the normalization standardization rule is a Min-Max standardization mode.

Preferably, in step S3, MSE is used as a loss function and the back propagation neural network is trained in an end-to-end manner, where the loss function is:

where m is the number of training data.

In order to achieve the above object, the present invention further provides an interactive multi-model system based on support vector regression, wherein the system comprises:

the model construction module is used for constructing an interactive multi-model algorithm based on support vector regression, and connecting a support vector regression network serving as a feedback network with a traditional interactive multi-model algorithm to obtain an adjusting coefficient of the algorithm;

the data preprocessing module is used for selecting a training set and a label set and preprocessing the training set and the label set according to a set normalization standard rule;

the network training module is used for training the support vector regression network model by using the training parameters according to the support vector regression network model and the training parameters thereof so as to finally achieve convergence; and;

and the estimation value calculation module is used for inputting the target state input interaction value and the measurement of the corresponding model to be processed into the improved model to obtain the target current time state estimation value.

To achieve the above object, the present invention further provides an interactive multi-model storage medium based on support vector regression, wherein the storage medium stores thereon a computer program, and the computer program, when executed by a processor, can implement the above steps of the interactive multi-model algorithm using support vector regression.

In order to achieve the above object, the present invention further provides an interactive multi-model computer device based on support vector regression, wherein the device comprises a memory, a processor and an algorithm program stored in the memory and operable on the processor, the algorithm program using the interactive multi-model algorithm based on support vector regression being configured to implement the steps of the interactive multi-model algorithm based on support vector regression

Compared with the prior art, the invention has the beneficial effects that:

the interactive multi-model method is based on the learning of the support vector regression network, and enhances the self-adaptive adjustment capability of the traditional interactive multi-model algorithm by means of the linear and nonlinear expression capability of the support vector regression network, wherein the support vector regression network is used as a feedback network to input interactive values and measures according to the target state corresponding to each model, firstly the interactive values and measures enter the feedback network to obtain coefficients for adjusting the process noise covariance of different models at the current moment, and finally the covariance coefficients are output to the traditional interactive multi-model algorithm to carry out current state estimation, so that the mapping between the model matching degree and the process noise covariance coefficients is established, and the tracking error when the models are not matched is greatly reduced.

Drawings

FIG. 1 is a flowchart of an interactive multi-model algorithm based on support vector regression according to the present invention;

FIG. 2 is a schematic structural diagram of an interactive multi-model system based on support vector regression according to the present invention;

FIG. 3 is a schematic block diagram of a support vector regression network in accordance with the present invention;

FIG. 4 is a schematic block diagram of an interactive multi-model algorithm based on support vector regression according to the present invention.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. 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 (b):

referring to fig. 1 to 4, the present invention provides a technical solution:

an interactive multi-model algorithm based on support vector regression, comprising the steps of:

and S1, constructing an interactive multi-model algorithm based on support vector regression, and connecting the support vector regression network serving as a feedback network with the traditional interactive multi-model algorithm to obtain the adjusting coefficient of the algorithm.

At each tracking moment, the target state input interaction value and the measurement corresponding to each model firstly enter a feedback network to obtain a coefficient for adjusting the process noise covariance of different models at the current moment, the coefficient is used as an adjusting coefficient, and the subsequent interactive multi-model algorithm estimates the target current state value according to the adjusting coefficient.

In addition, the number of input nodes of the vector regression network corresponds to the sum of the target state and the measurement dimension, the number of output layer nodes is 1, and the network relation is as follows:

and S2, selecting the training set and the label set, and preprocessing the training set and the label set according to a set normalization standard rule.

The training set comprises a plurality of target data sets with maneuvering grades, the target data sets are segmented to obtain the training set and a label set, the label set is the target real state at each moment, and the normalization standardization rule is a Min-Max standardization mode.

And S3, training the support vector regression network model by using the training parameters according to the support vector regression network model and the training parameters thereof, so as to finally achieve convergence.

The training parameters of the model are set, including learning rate and optimizer parameters, which can be adjusted according to training and testing effects, an optimal parameter value is finally selected, MSE is used as a loss function, and a back propagation neural network is trained in an end-to-end mode, wherein the loss function is as follows:

where m is the number of training data.

And S4, inputting the target state input interaction value and the measurement of the corresponding model to be processed into the improved model to obtain the target current time state estimation value.

In order to achieve the above object, the present invention further provides an interactive multi-model system based on support vector regression, wherein the system comprises:

the model construction module is used for constructing an interactive multi-model algorithm based on support vector regression, and connecting a support vector regression network serving as a feedback network with a traditional interactive multi-model algorithm to obtain an adjusting coefficient of the algorithm;

the data preprocessing module is used for selecting a training set and a label set and preprocessing the training set and the label set according to a set normalization standard rule;

the network training module is used for training the support vector regression network model by using the training parameters according to the support vector regression network model and the training parameters thereof so as to finally achieve convergence; and;

and the estimation value calculation module is used for inputting the target state input interaction value and the measurement of the corresponding model to be processed into the improved model to obtain the target current time state estimation value.

To achieve the above object, the present invention further provides an interactive multi-model storage medium based on support vector regression, wherein the storage medium stores thereon a computer program, and the computer program, when executed by a processor, can implement the above steps of the interactive multi-model algorithm using support vector regression.

The storage medium may include: any entity or device capable of carrying computer program code, recording medium, U.S. disk, removable hard disk, magnetic disk, optical disk, computer memory, Read Only Memory (ROM), Random Access Memory (RAM), electrical carrier wave signals, telecommunications signals, software distribution media, and the like. It should be noted that the computer readable medium may contain other components which may be suitably increased or decreased as required by legislation and patent practice in jurisdictions, for example, in some jurisdictions, in accordance with legislation and patent practice, the computer readable medium does not include electrical carrier signals and telecommunications signals.

The storage medium may include: any entity or device capable of carrying computer program code, recording medium, U.S. disk, removable hard disk, magnetic disk, optical disk, computer memory, Read Only Memory (ROM), Random Access Memory (RAM), electrical carrier wave signals, telecommunications signals, software distribution media, and the like. It should be noted that the computer readable medium may contain other components which may be suitably increased or decreased as required by legislation and patent practice in jurisdictions, for example, in some jurisdictions, in accordance with legislation and patent practice, the computer readable medium does not include electrical carrier signals and telecommunications signals.

In order to achieve the above object, the present invention further provides an interactive multi-model computer device based on support vector regression, wherein the device includes a memory, a processor, and an algorithm program stored in the memory and executable on the processor, the algorithm program using the interactive multi-model algorithm based on support vector regression being configured to implement the steps of the interactive multi-model algorithm based on support vector regression.

The computer equipment can be desktop computers, industrial computers, numerical control equipment, industrial robots, servers and other computing equipment. Those skilled in the art will appreciate that the computer device includes a processor and a memory, the description of which is for storing instructions is merely an example of a device and is not intended to limit the computer device, and may include more or fewer components, or some components in combination, or different components, e.g., the computer device may also include input-output devices, network access devices, buses, etc.

Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that changes, modifications, substitutions and alterations can be made in these embodiments without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.