CN112881835B - Electric characteristic sequence analysis-based battery car charging state analysis method - Google Patents

Abstract

The invention discloses a method for analyzing the charging state of an electric battery vehicle based on electrical characteristic sequence analysis, the method comprising the following steps: step S1, obtaining voltage and current sampling data when the electric battery vehicle is charging; step S2, setting a time interval, and dividing the voltage and current sampling data obtained in step S1 into a plurality of data sets; step S3, calculating the electrical characteristic quantity of each time interval; step S4, adding the calculation result of step S3 to a temperature sensor value, and generating an electrical characteristic sequence of the entire charging cycle; step S5, inputting the electrical characteristic sequence of the electric battery vehicle into a trained time recurrent neural network in sequence to obtain a result, if the result is 1, it indicates normal, and if it is 0, it indicates abnormal; compared with traditional temperature monitoring of electric battery vehicles, the invention uses electrical monitoring quantities to analyze the charging state of the electric battery vehicle, can quickly discover abnormal charging of the electric battery vehicle, and considers the time series changes of historical data and characteristics in the electrical characteristic sequence analysis, which can improve the accuracy of state analysis.

Description

A battery vehicle charging status analysis method based on electrical characteristic sequence analysis

Technical Field

The invention relates to a method for analyzing the charging state of an electric battery vehicle based on electrical characteristic sequence analysis, and belongs to the technical field of charging electric battery vehicles.

Background Art

With the popularity of electric vehicles, more and more people use them for travel. The uneven quality of batteries leads to frequent accidents in charging electric vehicles. How to realize the monitoring of the charging status of electric vehicles through technical means is a current hot issue. At present, no one has proposed an analysis method for the charging status of electric vehicles, and there is only a design method for charging devices for electric vehicles. For example, a Chinese patent application (publication number CN202586432U) proposes an electric vehicle monitoring and control system with charging protection, which includes a voltage and current detection unit, a Chinese patent application (publication number CN210866413U) proposes a device for monitoring and protecting the charging of electric vehicles, and the monitoring device can measure the deformation of the battery, and a Chinese patent application (publication number CN109742579B) proposes a fire prevention and anti-theft system for the charging socket of a battery vehicle, and a temperature sensor is added to the charging socket of the battery vehicle. The above Chinese patent applications can monitor the external data when the battery vehicle is charging, but they are unable to judge whether the charging status of the battery vehicle is normal or abnormal, such as common abnormalities such as battery leakage, charging short circuit, overvoltage charging, etc.

Summary of the invention

The technical problem to be solved by the present invention is: to provide a method for analyzing the charging status of an electric battery vehicle based on electrical feature sequence analysis. The method is based on a recurrent neural network method, analyzes the changes in various electrical features when the electric battery vehicle is charging, and then grasps the status of the electric battery vehicle when charging. It can detect abnormal conditions in time and prevent electrical fires caused by faults, so as to solve the technical problems existing in the prior art.

The technical solution adopted by the present invention is: a method for analyzing the charging state of a battery vehicle based on electrical characteristic sequence analysis, the method comprising the following steps:

Step S1, obtaining voltage and current sampling data when the battery vehicle is charging;

Step S2, setting a time interval, and dividing the voltage and current sampling data obtained in step S1 into several data sets;

Step S3, calculating the electrical characteristic quantity of each time interval;

Step S4, adding the calculation result of step S3 to the temperature sensing value T, and generating an electrical characteristic sequence of the entire charging cycle;

Step S5, input the electrical characteristic sequence of the battery vehicle into the trained time recurrent neural network in sequence to obtain a result. If the result is 1, it means that the charging state of the battery vehicle is normal. If the result is 0, it means that the charging state of the battery vehicle is abnormal.

The voltage and current sampling data acquired in step S1 are high-frequency original sampling data or low-frequency amplitude sampling data.

The voltage and current sampling data acquired in step S1 are replaced by current sampling data.

The electrical characteristic vector calculated in step S2 includes active power P, reactive power Q, apparent power S, power factor A combination of one or more of voltage harmonics U _i and current harmonics I _i , wherein voltage harmonics U _i and current harmonics I _i are obtained by fast Fourier transform (FFT) according to the voltage sampling data and current sampling data acquired in step S1, and voltage harmonics U _i and current harmonics I _i are in phasor form, and i is 0 to 11 or more.

The time recurrent neural network described in step S4 adopts a recurrent neural network RNN, a long short-term memory neural network LSTM, a gated recurrent unit GRU or a variant thereof.

Beneficial effects of the present invention: Compared with the prior art, compared with traditional battery vehicle temperature monitoring, the present invention uses electrical monitoring quantities to analyze the charging status of the battery vehicle, which can quickly discover abnormal charging of the battery vehicle, and takes historical data and time series changes into consideration when analyzing the characteristic sequence of electrical appliances, which can improve the accuracy of status analysis.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a flow chart of the present invention.

DETAILED DESCRIPTION

The present invention is further described below in conjunction with the accompanying drawings and specific embodiments.

Embodiment 1: As shown in FIG1 , a method for analyzing the charging state of a battery vehicle based on electrical characteristic sequence analysis comprises:

Step S1, collecting voltage and current data of the battery vehicle during charging through the smart socket, the sampling frequency of the smart socket is 6.4kHz/s, the sampling accuracy is 0.5 level, and the sampling time is from the start of charging to the completion of charging, which is assumed to be 7200s.

Step S2: Set a 1s time interval, and divide the voltage and current sampling data obtained in step S1 into 7200 data sets.

Step 3: Calculate the voltage sampling data and current sampling data of every 1s obtained in step S2 by fast Fourier transform algorithm (FFT) to obtain the 0-11th harmonics U _i of the voltage and the 0-11th harmonics I _i of the current, and calculate the active power P _i of each harmonic in turn:

P _i = U _i I _i

Where U _i is the i-th harmonic voltage amplitude, and I _i is the i-th harmonic current amplitude. The total active power P is obtained by summing the active powers of each harmonic:

Then, the apparent power S, reactive power Q and power factor cosφ are calculated based on the effective values of voltage and current.

S＝UI

Among them, U and I are the effective value of voltage and current respectively. Then the characteristics calculated every 1s are: active power P, apparent power S, reactive power Q and power factor cosφ, 0-11th harmonic voltage H _V , 0-11th harmonic current _HC .

Step S4: Add the calculated result of step S3 to the temperature sensor value T to construct the feature vector V

V＝[P,S,Q,cosφ,U ₁ ,...,U ₁₁ ,I ₁ ,...,I ₁₁ ,T]

The length of this vector is 29. Then the electrical characteristic sequence of the entire charging cycle can be expressed as:

[V ¹ ,V ² ,...,V ⁷²⁰⁰ ]

Where V ¹ represents the eigenvector of the 1st s, V ² represents the eigenvector of the 2nd s, and V ⁷²⁰⁰ represents the eigenvector of the 7200th s.

Step 5: Input the electrical characteristic sequence [V ¹ , V ² , ..., V ⁷²⁰⁰ ] of the battery vehicle into the trained time recurrent neural network in sequence to obtain the result [R ¹ , R ² , ..., R ⁷²⁰⁰ ]. If ^Ri = 1, it means that the charging state of the battery vehicle at the i-th second is normal. If ^Ri = 0, it means that the charging state of the battery vehicle at the i-th second is abnormal (battery leakage, charging short circuit, overvoltage charging, etc.).

Through the above steps, the charging status of the battery vehicle can be analyzed in real time according to the electrical monitoring quantity of the battery vehicle, abnormal charging status can be discovered in time, and the explosion and fire of the battery vehicle can be avoided.

The idea of the battery vehicle charging status analysis method based on electrical characteristic sequence analysis proposed in the present invention is that the battery vehicle charging status can be reflected by the voltage and current of the power supply port. Once the battery vehicle is in an abnormal state, it can be quickly reflected by the electrical monitoring quantity, and the problem can be quickly analyzed in combination with the recorded electrical characteristic sequence. However, existing monitoring methods such as temperature and deformation monitoring only react when the battery vehicle undergoes drastic changes such as high temperature and deformation. When the battery vehicle undergoes such reactions, it often enters a very dangerous state, and an explosion or electrical fire may occur at any time.

The above description is only a specific implementation mode of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims (4)

1. A method for analyzing the charging status of a battery vehicle based on electrical characteristic sequence analysis, characterized in that the method comprises the following steps: Step S1, obtaining voltage and current sampling data when the battery vehicle is charging; Step S2, setting a time interval, and dividing the voltage and current sampling data obtained in step S1 into several data sets; Step S3, calculating the electrical characteristic quantity of each time interval; The electrical characteristic vectors calculated in step S3 include active power P, reactive power Q, apparent power S, power factor A combination of one or more of voltage harmonics U _i and current harmonics I _i , wherein the voltage harmonics U _i and current harmonics I _i are obtained by fast Fourier transform (FFT) according to the voltage sampling data and current sampling data acquired in step S1, and the voltage harmonics U _i and current harmonics I _i are in phasor form, and i ranges from 0 to 11; active power P _i : P _i = U _i I _i Where _Ui is the i-th harmonic voltage amplitude, _Ii is the i-th harmonic current amplitude, and the total active power P is obtained by summing up the active powers of each harmonic: Then, the apparent power S, reactive power Q and power factor cosφ are calculated based on the effective values of voltage and current; S＝UI Where U and I are the effective value of voltage and current respectively, then the characteristics calculated every 1s are: active power P, apparent power S, reactive power Q and power factor cosφ, 0-11th harmonic voltage H _V , 0-11th harmonic current _HC ; Step S4: Add the calculated result of step S3 to the temperature sensor value T, and generate the electrical characteristic sequence of the entire charging cycle, that is, construct the characteristic vector V V＝[P,S,Q,cosφ,U ₁ ,...,U ₁₁ ,I ₁ ,...,I ₁₁ ,T] The length of this vector is 29, so the electrical characteristic sequence of the entire charging cycle can be expressed as: [V ¹ ,V ² ,...,V ⁷²⁰⁰ ] Where V ¹ represents the eigenvector of the 1st s, V ² represents the eigenvector of the 2nd s, and V ⁷²⁰⁰ represents the eigenvector of the 7200th s; Step S5: input the electrical characteristic sequence [V ¹ , V ² , ..., V ⁷²⁰⁰ ] of the battery vehicle into the trained time recurrent neural network in sequence to obtain the result [R ¹ , R ² , ..., R ⁷²⁰⁰ ]. If the result ^Ri is 1, it means that the charging state of the battery vehicle is normal; if the result ^Ri is 0, it means that the charging state of the battery vehicle is abnormal. 2. A battery vehicle charging state analysis method based on electrical characteristic sequence analysis according to claim 1, characterized in that: the voltage and current sampling data acquired in step S1 are high-frequency original sampling data or low-frequency amplitude sampling data. 3. The method for analyzing the charging status of a battery vehicle based on electrical characteristic sequence analysis according to claim 1 is characterized in that: the voltage and current sampling data obtained in step S1 are replaced by current sampling data. 4. A battery vehicle charging status analysis method based on electrical feature sequence analysis according to claim 1, characterized in that: the time recurrent neural network described in step S4 adopts a recurrent neural network RNN, a long short-term memory neural network LSTM, a gated recurrent unit GRU or a variant method thereof.