CN111090048B - A new energy vehicle on-board data adaptive time interval transmission method - Google Patents

Abstract

The invention relates to a method for self-adapting time interval transmission of on-board data of a new energy vehicle, and belongs to the field of on-board data processing. The method includes: S1 selecting the dynamic working condition data of the power battery under laboratory conditions or actual energy vehicles, collecting and arranging the technical parameters of the battery; S2 intercepting a section of voltage, temperature or current data, using Haar wavelet transform to extract the wavelet decomposition coefficient, After coefficient processing, wavelet reconstruction is performed; S3 records the initial moment of each segment and the corresponding original voltage, temperature or current data according to the reconstructed voltage, temperature or current data, and obtains the voltage, temperature or current data after dimension reduction, And record other on-board data at the corresponding moment; S4 models and estimates the state of the processed battery data. The invention can obtain on-board data for adaptive time interval transmission, ensure the integrity of data under severe working conditions, and has higher modeling and state estimation accuracy, and has more advantages than the current fixed time interval transmission.

Description

A new energy vehicle on-board data adaptive time interval transmission method

technical field

The invention belongs to the field of on-board data processing, and relates to a method for self-adapting time interval transmission of on-board data of new energy vehicles.

Background technique

With the advent of the era of the Internet of Things and intelligent automobiles, a large amount of vehicle data is connected to the cloud data center through the Internet of Things, and various types of data uploaded in real time and at high frequency by the vehicle intelligent terminal will be collected in the data center. Mass data. For the collection, cleaning, conversion, storage, real-time and offline data analysis and value mining of these data, major auto companies have built their own big data platforms to realize information such as battery status, vehicle status and geographic location of new energy vehicles. 24/7 real-time monitoring.

At present, the data communication between the vehicle terminal, vehicle enterprise platform and public platform of new energy vehicles follows the national standard "Technical Specification for Electric Vehicle Remote Service and Management System" (GT32960-2016). Among them, the standard states that the transmission time interval of real-time data uploaded from the vehicle terminal to the enterprise platform should not exceed 30s at most. It is understood that the time interval adopted by the big data platforms of various enterprises is mostly fixed 10s, which will lead to the data uploaded to the platform may be missing under severe working conditions or short-term changes in working conditions, thus missing some useful information. value data. Therefore, there is a need for a transmission method that can adapt to the time interval, so that it can transmit more data points under severe conditions and fewer data points under moderate conditions.

SUMMARY OF THE INVENTION

In view of this, the purpose of the present invention is to provide an adaptive time interval transmission method for on-board data of new energy vehicles, which realizes the adaptive interval data transmission of on-board data and ensures the accuracy of big data analysis results.

To achieve the above object, the present invention provides the following technical solutions:

A method for self-adaptive time interval transmission of on-board data of a new energy vehicle, which specifically includes the following steps:

S1: Select the dynamic working condition data of the power battery under laboratory conditions or actual energy vehicles, and collect and sort out the technical parameters of the battery;

S2: Intercept a section of vehicle battery data such as voltage, temperature or current (selected according to your own needs), use Haar wavelet transform to extract wavelet decomposition coefficients, and then perform wavelet reconstruction after processing the coefficients;

S3: Record the initial moment of each segment and the corresponding original voltage, temperature or current data according to the reconstructed voltage, temperature or current data, obtain the voltage, temperature or current data after dimension reduction, and record the other on-board batteries at the corresponding moment data;

S4: Model and estimate the state of the processed battery data, and compare the estimation accuracy of the current fixed interval processing method.

Further, the step S2 specifically includes the following steps:

S21: For the intercepted voltage, temperature or current data, the data length must meet the m power of 2, if not, then fill it with 0;

S22: Perform Haar wavelet decomposition on the voltage, temperature or current data, and arrange the wavelet coefficients in descending order;

S23: Select the threshold δ, and set all the sorted wavelet coefficients less than δ to 0, that is, set the values of the original wavelet coefficients less than the threshold to 0;

S24: Integrate the wavelet coefficients at this time and then perform wavelet reconstruction to obtain a series of piecewise constant data.

Further, the step S3 specifically includes the following steps:

S31: If the length of the reconstructed data exceeds the original length N, truncate from N;

S32: Record the initial moment corresponding to the first point of each segment, find the voltage, temperature or current value of the original data at this moment, form a new set of voltage, temperature or current data, and record other vehicle-mounted data at the corresponding moment.

Further, the step S4 specifically includes the following steps:

S41: establish a battery model, such as a first-order equivalent circuit model;

S42: select an appropriate parameter identification and state estimation algorithm;

S43: Use the new adaptive interval battery data and the battery data extracted at the fixed interval to train the model respectively, and obtain the accuracy of the comparative state estimation.

Further, in the step S41, establishing a thermal model is also included.

Further, in the step S42, the state estimation algorithm adopted is specifically: joint estimation using recursive least squares (Recursive Least Square, RLS) and Extended Kalman Filter (Extended Kalman Filter, EKF), but not limited to this algorithm.

Further, in the step S43, the state estimation mainly specifically is to perform battery state of charge (State of Charge, SOC) estimation.

The beneficial effects of the present invention are:

1) The self-adaptive time interval transmission method adopted in the present invention is to automatically extract the on-board data according to the driving state of the vehicle, so that the on-board terminal can transmit more data points under severe working conditions, and transmit less data under gentle working conditions. point;

2) The present invention mainly applies the characteristics of Haar wavelet transform, the algorithm is simple, and has obvious applicability and feasibility;

3) The vehicle-mounted data obtained according to the present invention is more suitable for the actual data, and has higher accuracy in the later data analysis and data mining.

Other advantages, objects, and features of the present invention will be set forth in the description that follows, and will be apparent to those skilled in the art based on a study of the following, to the extent that is taught in the practice of the present invention. The objectives and other advantages of the present invention may be realized and attained by the following description.

Description of drawings

In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be preferably described in detail below with reference to the accompanying drawings, wherein:

Fig. 1 is the realization flow chart of a kind of new energy vehicle in-vehicle data adaptive time interval transmission method according to the present invention;

Figure 2 is a comparison between the reconstructed voltage data and the original data;

Figure 3 is a first-order equivalent circuit model;

Fig. 4 is the data and SOC estimation result of the adaptive time interval;

Figure 5 shows the data and SOC estimation results for a fixed time interval of 10s.

Detailed ways

The embodiments of the present invention are described below through specific specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the drawings provided in the following embodiments are only used to illustrate the basic idea of the present invention in a schematic manner, and the following embodiments and features in the embodiments can be combined with each other without conflict.

Please refer to FIG. 1 to FIG. 3. FIG. 1 is a preferred method for transmitting on-board data of a new energy vehicle at an adaptive time interval, which specifically includes the following steps:

S1: Select the dynamic working condition data of the power battery under laboratory conditions or in a real vehicle. In this example, a set of charge and discharge data of the ternary prismatic battery of AVIC Lithium Battery under laboratory conditions is selected, and the technical parameters of the battery are collected and sorted;

S2: Intercept a section of voltage data (or other vehicle-mounted data, such as temperature, current, etc., selected according to your own needs), use Haar wavelet transform to extract wavelet decomposition coefficients, and perform wavelet reconstruction after processing the coefficients. Specifically include the following steps:

S21: For the intercepted voltage data U _origin , the data length is recorded as N, and if N is not equal to the k power of 2, it is filled with 0;

S22: Perform Haar wavelet decomposition on the voltage data, arrange the wavelet coefficients in descending order, and call the Matlab function waverec( ) to realize the wavelet transform;

[C,L]=wavedec(U _origin ,3,'haar')

Among them, C is the wavelet coefficient, C=[C ₁ , C ₂ ,...,C _i ,...,Cn], n=2 ^k , L is the number of the corresponding wavelet coefficients, and haar means using Haar wavelet decomposition.

S23: Select the threshold δ, and set all the sorted wavelet coefficients less than δ to 0, that is, set the values of the original wavelet coefficients less than the threshold to 0;

f(C _i <δ), C _i =0

The wavelet coefficient at this time is recorded as C _rec .

S24: Integrate the wavelet coefficients at this time and then perform wavelet reconstruction to obtain a series of piecewise constant data.

U _rec =waverec(C _rec ,L,'haar')

S3: Record the initial time of each segment and the corresponding original voltage data according to the reconstructed voltage data, obtain the voltage data after dimension reduction, and record other vehicle-mounted data at the corresponding time. Specifically include the following steps:

S31: If the length of the reconstructed data exceeds the original length N, truncate from N, and Figure 2 shows the comparison between the reconstructed voltage data and the original data;

S32: Record the initial moment corresponding to the first point of each segment, find the voltage value of the original data at this moment, form a new set of voltage data, and record other vehicle-mounted data at the corresponding moment;

A=[(t ₁ ,U ₁ ),(t ₂ ,U ₂ ),…,(t _i ,U _i ),…,(t _m ,U _m )]

Among them, m is the number of subsequences, t _i is the initial moment of each subsequence, and U _i is the voltage data in U _origin corresponding to t _i .

Then the voltage sequence at this time is U _a =[U ₁ ,U ₂ ,...,U _i ,...,U _m ]

S4: Modeling and state estimation are performed on the processed battery data, and the estimation accuracy of the current fixed time interval processing method is compared. Specifically include the following steps:

S41: establish a battery model, and select a first-order equivalent circuit model in this example, as shown in Figure 3;

S42: Select an appropriate parameter identification and state estimation algorithm. Here, RLS and EKF are used for joint estimation. The estimation process is as follows:

Q,R，其中，

分别为RLS的初始参数值和参数估计的误差协方差矩阵初始值；

Q,R分别为EKF的状态初始值、状态估计误差协方差矩阵初始值、系统噪声协方差矩阵和观测噪声协方差。Parameter initialization:

Q, R, where,

are the initial parameter value of RLS and the initial value of the error covariance matrix of parameter estimation;

Q, R are the initial value of the state of the EKF, the initial value of the state estimation error covariance matrix, the system noise covariance matrix and the observation noise covariance, respectively.

Time updates of state variables:

in,

Parameter estimation for RLS:

Among them, λ is the forgetting factor, y _k =U _t,k -OCV _k .

State updates for state variables:

in,

S43: Train the model by using the battery data of the new adaptive time interval and the battery data extracted at the fixed interval, and compare the accuracy of the state estimation.

Figure 4 shows the data and the SOC estimation result of the adaptive time interval using the method of the present invention, and Figure 5 shows the data and the SOC estimation result of the fixed 10s time interval. Comparing Figures 4 and 5, it can be seen that the method of the present invention estimates the SOC The error is within 1%, while the SOC estimation error is within 2% using the fixed time interval data of 10s. Obviously, the adaptive time interval method proposed by the present invention has more advantages.

Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be Modifications or equivalent replacements, without departing from the spirit and scope of the technical solution, should all be included in the scope of the claims of the present invention.

Claims (7)

1. A new energy automobile vehicle-mounted data self-adaptive time interval transmission method is characterized by comprising the following steps:

s1: selecting dynamic working condition data of the power battery of the actual energy automobile under laboratory conditions, and collecting technical parameters of the battery;

s2: intercepting a section of voltage, temperature or current data, extracting a wavelet decomposition coefficient by using haar wavelet transform, and performing wavelet reconstruction after processing the coefficient;

s3: recording the initial time of each section and the corresponding original voltage, temperature or current data according to the reconstructed voltage, temperature or current data to obtain the reduced-dimension voltage, temperature or current data, and recording other vehicle-mounted battery data at the corresponding time;

s4: and modeling and state estimation are carried out on the processed battery data.

2. The adaptive time interval transmission method for the vehicle-mounted data of the new energy vehicle as claimed in claim 1, wherein the step S2 specifically includes the following steps:

s21: for the intercepted voltage, temperature or current data, the data length needs to satisfy the m power of 2, if not, the data length is complemented by 0;

s22: carrying out haar wavelet decomposition on the voltage, temperature or current data, and carrying out descending arrangement on wavelet coefficients;

s23: selecting a threshold value delta, setting all the sorted wavelet coefficients smaller than the delta to be 0, namely setting the numerical value of the original wavelet coefficient smaller than the threshold value to be 0;

s24: and integrating the wavelet coefficients at the moment, and performing wavelet reconstruction to obtain a series of piecewise constant data.

3. The adaptive time interval transmission method for the vehicle-mounted data of the new energy vehicle as claimed in claim 1, wherein the step S3 specifically includes the following steps:

s31: if the reconstructed data length exceeds the original length N, truncating from N;

s32: and recording the initial moment corresponding to each section of the first point, finding the voltage, temperature or current value of the original data at the moment to form a new group of voltage, temperature or current data, and recording other vehicle-mounted data at the corresponding moment.

4. The adaptive time interval transmission method for the vehicle-mounted data of the new energy vehicle as claimed in claim 1, wherein the step S4 specifically includes the following steps:

s41: establishing a battery model;

s42: selecting a proper parameter identification and state estimation algorithm;

s43: and training the model by using the new self-adaptive interval battery data to obtain state estimation.

5. The adaptive time interval transmission method for the vehicle-mounted data of the new energy vehicle according to claim 4, wherein the step S41 further comprises establishing a thermal model.

6. The adaptive time interval transmission method for the vehicle-mounted data of the new energy vehicle according to claim 4, wherein in the step S42, the adopted state estimation algorithm specifically comprises: joint estimation using Recursive Least Squares (RLS) and Extended Kalman Filter (EKF) is used.

7. The on-board data adaptive time interval transmission method for the new energy vehicle according to claim 4, wherein in the step S43, the State estimation is specifically a State of Charge (SOC) estimation.

Images