1. The on-line estimator of the charge state of the lithium ion battery with the measurement data anomaly detection function is characterized by comprising an information acquisition unit, an anomaly data detection unit, an information processing unit, a control unit and a display unit;

the information acquisition unit is used for acquiring the voltage, the current and the temperature of the lithium battery or the lithium battery pack;

the abnormal data detection unit is used for receiving the data of the information acquisition unit and sending the processed data to the information processing unit;

the control unit is used for controlling the data transmission between the information acquisition unit and the information processing unit and monitoring the work of the estimator;

the display unit is connected with the information processing unit.

2. The lithium ion battery state of charge on-line estimator with measurement data anomaly detection of claim 1, wherein:

the control unit controls the transmission process of the information acquisition unit and the information processing unit and monitors whether the state online estimator is in a normal environment or not;

the abnormal data detection unit detects whether the acquired data is in a normal range or not, and corrects the data when the data is abnormal;

and the information processing unit receives the data from the abnormality detection unit, estimates the state of the lithium battery according to an improved particle filter algorithm, and stores and displays the estimation result in real time.

3. The lithium ion battery state of charge on-line estimator with measurement data anomaly detection of claim 2, wherein: the information processing unit identifies equivalent circuit parameters by using a recursive least square algorithm with a rectangular window, and estimates the state of the lithium battery by combining a genetic algorithm and particle filtering.

4. The lithium ion battery state of charge on-line estimator with measurement data anomaly detection of claim 2, wherein: the abnormal data detection unit multiplies the current measured at the moment by the internal resistance range of the lithium battery obtained in advance to obtain a limit range, if the voltage measured at the moment and the voltage at the last moment exceed the range, the abnormal data detection unit judges that the data are abnormal, discards the voltage data at the moment and takes the data at the last moment as the value at the moment.

5. The lithium ion battery state of charge on-line estimator with measurement data anomaly detection of claim 2, wherein: the improved particle filtering algorithm comprises the following processes: after generating the initial particles, the following loop is performed until the prediction is finished: updating the particle position, updating the particle weight, calculating an estimated value, and judging whether to resample: if so, performing genetic algorithm resampling and predicting the next moment; if not, directly predicting the next moment, judging whether to finish the prediction, and if not, returning to the step of updating the particle position;

the genetic algorithm resampling includes intersection and variation of particles.

6. The online estimation method for the state of charge of the lithium ion battery with the measurement data anomaly detection function is characterized by comprising the following steps of:

step S1: obtaining the voltage, the current and the temperature value of the lithium battery during working by using a sensor;

step S2: carrying out data anomaly detection processing on the acquired data;

step S3: performing online identification on the equivalent circuit parameters of the lithium battery by using a recursive least square algorithm;

step S4: and establishing a state equation of the equivalent circuit model, and estimating the state of the battery by using an improved particle filter algorithm.

7. The online estimation method for the state of charge of the lithium ion battery with measurement data anomaly detection according to claim 6, characterized in that: in step S2, the specific method of data anomaly detection processing is: multiplying the current data at the current moment by the internal resistance range estimated in advance, subtracting the voltage at the current moment from the voltage at the previous moment, and comparing the two to obtain a judgment condition: if the latter data is smaller than the former, the data is valid; otherwise, the voltage data at the current moment is discarded and is filled by the voltage data at the previous moment.

8. The online estimation method for the state of charge of the lithium ion battery with measurement data anomaly detection according to claim 6, characterized in that: in step S3, the lithium battery equivalent circuit is a second-order Thevenin equivalent circuit.

9. The online estimation method for the state of charge of the lithium ion battery with measurement data anomaly detection according to claim 6, characterized in that: in step S4, the specific steps of the improved particle filtering algorithm are as follows:

step S41: generating initial particles N according to the prior probability, wherein the initial weight of all the particles is 1/N;

step S42: updating the particle position and the weight value, and calculating an estimated value at the moment;

step S43: judging whether resampling is needed, if not, returning to the step S42 for estimating the next moment; and if resampling is needed, carrying out the operations of crossing and mutation of the particles to obtain a new particle set.

10. The online estimation method for the state of charge of the lithium ion battery with measurement data anomaly detection according to claim 9, characterized in that: in step S43, a specific method of determining whether resampling is necessary is to determine based on the particle variance at that time.

Background

The lithium battery state of charge estimation method can be roughly divided into a traditional algorithm and an intelligent algorithm, wherein the traditional algorithm comprises an open-circuit voltage method, an internal resistance method, an ampere-hour integral method and the like, and the lithium battery state of charge estimation method has the advantages of being simple in calculation and large in error, difficult to measure, not suitable for real-time estimation and the like. The traditional algorithms have more limitations, so the current research focus is on intelligent algorithms, including neural network algorithms, kalman filtering, particle filtering, and the like. The neural network algorithm needs a large amount of data support, and the accuracy of the Kalman filtering in battery estimation is influenced by the applicable condition of the Kalman filtering.

Disclosure of Invention

In view of the above, in order to make up for the blank and the deficiency of the prior art, the present invention aims to provide an online estimator and method for a state of charge of a lithium ion battery with measurement data anomaly detection, so as to reduce the problems of increased calculation amount and reduced estimation accuracy caused by measurement data anomaly.

The system comprises an information acquisition unit, a control unit, a data abnormity detection unit, an information processing unit and a display unit. The method comprises the steps that the information acquisition unit acquires the terminal voltage of the lithium battery, the temperature of the lithium battery and the current information flowing through the lithium battery, the control unit controls the transmission process of the acquisition unit and the information processing unit and monitors whether the state estimator is in a normal environment or not, the data abnormity detection unit detects whether acquired data are in a normal range or not, data correction is carried out according to a certain rule when the data are abnormal, the information processing unit estimates the state of the lithium battery according to an improved particle filter algorithm and stores and displays the estimated result in real time according to the received data from the abnormity detection unit. The state estimator identifies equivalent circuit parameters by using a recursive least square algorithm with a rectangular window, estimates the state of the lithium battery by combining a genetic algorithm and particle filtering, solves the problem of particle depletion of the particle filtering, improves estimation precision, and ensures that the abnormal data does not increase the calculated amount of the particle filtering and ensures the calculation precision by using an abnormal data detection unit.

The invention specifically adopts the following technical scheme:

the on-line estimator of the charge state of the lithium ion battery with the measurement data anomaly detection function is characterized by comprising an information acquisition unit, an anomaly data detection unit, an information processing unit, a control unit and a display unit;

the information acquisition unit is used for acquiring the voltage, the current and the temperature of the lithium battery or the lithium battery pack;

the abnormal data detection unit is used for receiving the data of the information acquisition unit and sending the processed data to the information processing unit;

the control unit is used for controlling the data transmission between the information acquisition unit and the information processing unit and monitoring the work of the estimator;

the display unit is connected with the information processing unit.

Further, the control unit controls the transmission process of the information acquisition unit and the information processing unit and monitors whether the state online estimator is in a normal environment;

the abnormal data detection unit detects whether the acquired data is in a normal range or not, and corrects the data when the data is abnormal;

and the information processing unit receives the data from the abnormality detection unit, estimates the state of the lithium battery according to an improved particle filter algorithm, and stores and displays the estimation result in real time.

Further, the information processing unit identifies equivalent circuit parameters by using a recursive least square algorithm with a rectangular window, and estimates the state of the lithium battery by combining a genetic algorithm and particle filtering.

Further, the abnormal data detection unit multiplies the current measured at the moment by the internal resistance range of the lithium battery obtained in advance to obtain a limit range, if the voltage measured at the moment and the voltage at the last moment exceed the limit range, the abnormal data detection unit judges that the data is abnormal, discards the voltage data at the moment and takes the data at the last moment as the value at the moment.

Further, the improved particle filtering algorithm comprises the following processes: after generating the initial particles, the following loop is performed until the prediction is finished: updating the particle position, updating the particle weight, calculating an estimated value, and judging whether to resample: if so, performing genetic algorithm resampling and predicting the next moment; if not, directly predicting the next moment, judging whether to finish the prediction, and if not, returning to the step of updating the particle position;

the genetic algorithm resampling includes intersection and variation of particles.

The online estimation method for the state of charge of the lithium ion battery with the measurement data anomaly detection is characterized by comprising the following steps of:

step S1: obtaining the voltage, the current and the temperature value of the lithium battery during working by using a sensor;

step S2: carrying out data anomaly detection processing on the acquired data;

step S3: performing online identification on the equivalent circuit parameters of the lithium battery by using a recursive least square algorithm;

step S4: and establishing a state equation of the equivalent circuit model, and estimating the state of the battery by using an improved particle filter algorithm.

Further, in step S2, the specific method of the data anomaly detection processing is as follows: multiplying the current data at the current moment by the internal resistance range estimated in advance, subtracting the voltage at the current moment from the voltage at the previous moment, and comparing the two to obtain a judgment condition: if the latter data is smaller than the former, the data is valid; otherwise, the voltage data at the current moment is discarded and is filled by the voltage data at the previous moment.

Further, in step S3, the lithium battery equivalent circuit is a second-order Thevenin equivalent circuit.

Further, in step S4, the specific steps of the improved particle filtering algorithm are as follows:

step S41: generating initial particles N according to the prior probability, wherein the initial weight of all the particles is 1/N;

step S42: updating the particle position and the weight value, and calculating an estimated value at the moment;

step S43: judging whether resampling is needed, if not, returning to the step S42 for estimating the next moment; and if resampling is needed, carrying out the operations of crossing and mutation of the particles to obtain a new particle set.

Further, in step S43, a specific method of determining whether resampling is required is to determine based on the particle variance at that time.

Compared with the prior art, the method and the device solve the problems of low precision and large calculation amount caused by measurement data loss during online estimation of the state of charge of the lithium ion battery. The method comprises the steps that the information acquisition unit acquires the terminal voltage of the lithium battery, the temperature of the lithium battery and the current information flowing through the lithium battery, the control unit controls the transmission process of the acquisition unit and the information processing unit and monitors whether the state estimator is in a normal environment or not, the data abnormity detection unit detects whether the acquired data is in a normal range or not, data correction is carried out according to a certain rule when the data is abnormal, the information processing unit estimates the state of the lithium battery according to an improved particle filtering algorithm according to the data received from the abnormity detection unit, and the estimated result is stored and displayed in real time. The state estimator identifies equivalent circuit parameters by using a recursive least square algorithm with a rectangular window, estimates the state of the lithium battery by combining a genetic algorithm and particle filtering, solves the problem of particle depletion of the particle filtering, improves estimation precision, and ensures that the abnormal data does not increase the calculated amount of the particle filtering and ensures the calculation precision by using an abnormal data detection unit.

Drawings

The invention is described in further detail below with reference to the following figures and detailed description:

FIG. 1 is a schematic structural diagram of an apparatus according to an embodiment of the present invention;

fig. 2 is a schematic overall flow chart of an online estimation method for the state of charge of a lithium ion battery with abnormal measurement data according to an embodiment of the present invention;

FIG. 3 is a schematic diagram illustrating a process of identifying parameters of an equivalent circuit model of a lithium battery according to an embodiment of the present invention;

FIG. 4 is a flowchart illustrating exception handling for metrology data according to an embodiment of the present invention;

FIG. 5 is a schematic flow chart of an improved particle filter algorithm according to an embodiment of the present invention;

fig. 6 is a schematic diagram of a second-order Thevenin equivalent circuit according to an embodiment of the present invention.

Detailed Description

In order to make the features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in detail as follows:

referring to fig. 1, the present embodiment provides a design scheme of an online state-of-charge estimator for a lithium ion battery with measurement data anomaly detection, including: the system comprises an information acquisition unit, an abnormal data detection unit, an information processing unit, a control unit and a display unit; the information acquisition unit acquires the voltage, the current and the temperature of the lithium battery or the lithium battery pack; the abnormal data detection unit receives the data of the information acquisition unit, and the processed data is sent to the information processing unit; the control unit controls the data transmission between the information acquisition unit and the information processing unit and monitors the work of the estimator; the display unit is connected with the information processing unit. As shown in fig. 1, the information acquisition unit acquires and transmits information such as voltage, current, temperature and the like of the lithium battery to the abnormal data detection unit, the abnormal data detection unit judges and processes the data, the information processing unit performs state estimation according to the estimation method, and finally, required information is transmitted to the upper computer and the display unit.

The invention provides an online estimation method for the state of charge of a lithium ion battery of a measurement data anomaly detection device, the flow is shown in figure 2, and the online estimation method is characterized in that: the method comprises the following steps:

step S1: the acquisition unit is a series of sensors and is used for acquiring the terminal voltage of the lithium battery, the current flowing through the lithium battery and the temperature information of the lithium battery.

Step S2: before the collected data are transmitted to the information processing unit for processing, the abnormal data detection unit judges whether the data are in the fluctuation range. The judgment flow chart is shown in fig. 4, namely: and multiplying the current data at the current moment by the internal resistance range estimated in advance, subtracting the voltage at the current moment from the voltage at the previous moment, and comparing the two to obtain a judgment condition. If the latter data is smaller than the former data, the data is valid, otherwise, the voltage data at the current moment is abandoned and filled by the voltage data at the previous moment.

Step S3: the parameters of the lithium battery equivalent circuit are identified on line by using a recursive least square algorithm, and the lithium battery equivalent circuit is a second-order Thevenin equivalent circuit as shown in fig. 6.

Referring to fig. 3, firstly, a frequency domain equation (1) of the circuit is established, and the discrete transformation is in the form of (2), so as to obtain an equation in the form of:

in the formula tau₁、τ₂Is the time constant of the RC element, i.e. tau₁＝R₁×C₁，τ₂＝R₂×C₂(ii) a Coefficient a_k(k＝1，2)、b_k(k ═ 1, 2, 3) is an unknown coefficient which is equivalent to the parameter R in the circuit₀、R₁、R₂、C₁、C₂Has a certain mathematical relationship and can be calculated by the following process.

Bilinear inverse transformation formulaSubstituting the formula (2), comparing the obtained formula with the formula (1) to obtain the relationship between the battery parameters and the coefficients, and calculating according to the flow of FIG. 3 to obtain the battery parameters and coefficientsAnd the coefficient of each moment is used for obtaining the equivalent circuit model parameter according to the relational expression. In the formula: t is the sampling time.

Step S4: establishing a state equation of an equivalent circuit model, and estimating the state of the battery by using an improved particle filter algorithm:

it can first establish the discrete state equation of the system:

in the formula: t is sampling time; u shape_p1、U_p2Voltages for the RC link, i.e. corresponding to U in the circuit₁、U₂(ii) a k is a certain current moment, and k +1 is the current next moment; omega_k、υ_kRespectively process noise and observation noise at the time k; h is_SOC-OCVIs a function relation of open-circuit voltage and state of charge (SOC); s (k +1) is the battery state of charge SOC at time k + 1.

According to the improved particle filter algorithm flowchart, real-time estimation of the battery state of charge is performed, as shown in fig. 5, which includes the following steps:

step S41: generating initial particles N according to the prior probability, wherein the initial weight of all the particles is 1/N;

step S42: updating the particle position and the weight value, and calculating an estimated value at the moment;

step S43: and judging whether resampling is needed or not, and judging according to the particle variance at the moment. If resampling is not needed, returning to the step S42 to estimate the next moment; and if resampling is needed, carrying out the operations of crossing and mutation of the particles to obtain a new particle set.

The present invention is not limited to the above preferred embodiments, and any other various types of on-line estimation methods and apparatuses for lithium ion battery state of charge with measurement data anomaly detection can be obtained from the present invention, and all equivalent changes and modifications made according to the claimed scope of the present invention shall fall within the scope of the present invention.