Large-scale electric heavy truck charging and discharging cooperative control method and system

Abstract

The invention discloses a large-scale electric heavy truck charging and discharging cooperative control method and system, and the method comprises the steps: collecting vehicle operation data of an electric heavy truck in an operation state, and extracting the feature data of the vehicle operation data of the electric heavy truck; based on the extracted historical vehicle operation characteristic data of the electric heavy truck, constructing a prediction model of the operation data of the electric heavy truck, and obtaining a vehicle operation characteristic data prediction value of the electric heavy truck; and setting a charge and discharge demand priority index based on the vehicle operation characteristic data predicted values of different electric heavy trucks, and calculating the charge and discharge power of the electric heavy trucks based on the charge and discharge demand priority index. According to the method, the operation data of the electric heavy truck are collected, the feature data are extracted, the prediction model of the operation data of the electric heavy truck is constructed, and the refined charging and discharging cooperative control of the electric heavy truck is realized through the charging and discharging demand priority index and the dynamic power distribution method.

Description

Technical Field

[0001] The present invention belongs to the technical field of electric heavy truck charge-discharge collaborative control, and particularly relates to a method and system for large-scale electric heavy truck charge-discharge collaborative control. Background Art

[0002] With the advancement of environmental protection policies and the progress of new energy technologies, the application of electric heavy trucks in scenarios such as logistics, mines, and ports has gradually become popular. Electric heavy trucks have advantages such as zero emissions, low noise, and low operating costs. However, their large-scale application also brings new challenges. Electric heavy trucks have large battery capacities, high charging requirements, and complex operating conditions (such as frequent starts and stops, large load changes, etc.), which pose higher requirements for battery management and charge-discharge control. When electric heavy trucks are applied on a large scale, simultaneous charging of a large number of vehicles will cause huge pressure on the power grid, which may lead to grid overload or voltage instability;

[0003] The current charge-discharge collaborative technology for electric heavy trucks is mainly based on fixed rules (such as timed charging, charging when the battery level is below a threshold, etc.), and it is difficult to adapt to the complex and changeable operating conditions of electric heavy trucks. Therefore, there is an urgent need for a method for large-scale electric heavy truck charge-discharge collaborative control to solve the technical problems existing in the current charge-discharge collaboration of electric heavy trucks. Summary of the Invention

[0004] In order to solve the deficiencies in the prior art, the purpose of the present invention is to provide a method and system for large-scale electric heavy truck charge-discharge collaborative control, which is used to solve the technical problem that simultaneous charging of a large number of vehicles in the prior art will cause huge pressure on the power grid, which may lead to grid overload or voltage instability.

[0005] The present invention adopts the following technical solutions.

[0006] The present invention proposes a method for large-scale electric heavy truck charge-discharge collaborative control, including:

[0007] Step 1: Collect vehicle operation data of electric heavy trucks under operating conditions, and extract characteristic data of the vehicle operation data of electric heavy trucks;

[0008] Step 2: Based on the extracted historical vehicle operation characteristic data of electric heavy trucks, construct a prediction model of the electric heavy truck operation data, and obtain the predicted values of the vehicle operation characteristic data of electric heavy trucks;

[0009] Step 3: Based on the predicted values of the vehicle operation characteristic data of different electric heavy trucks, set the charge-discharge demand priority index, and calculate the charge-discharge power of electric heavy trucks based on the charge-discharge demand priority index.

[0010] Furthermore, the method for collecting vehicle operation data of electric heavy trucks under operating conditions and extracting characteristic data of the vehicle operation data of electric heavy trucks is as follows:

[0011] Dynamically divide time intervals, where the length of each time interval is T, and vehicle operation data is collected in real time by on-vehicle sensors within each time interval; the vehicle operation data includes battery status data, vehicle driving data, and vehicle load data; define the time when the absolute value of the acceleration of the electric heavy truck during driving is greater than or equal to a as the variable acceleration driving time, the mileage traveled by the electric heavy truck during the variable acceleration driving duration as the variable acceleration driving mileage, and a continuous period of variable acceleration driving time as one variable acceleration driving; define the time period during which the load weight of the electric heavy truck remains unchanged during driving as a load status time period;

[0012] Integrate the remaining power data, real-time current data, temperature data, and internal resistance data in the battery status data to extract battery status features;

[0013] Integrate the real-time speed of the electric heavy truck, variable acceleration driving duration, and variable acceleration driving mileage in the vehicle driving status data to extract vehicle driving features;

[0014] Integrate the real-time load weight of the electric heavy truck and load driving data in the vehicle load status data to extract vehicle load features.

[0015] Furthermore, integrate the remaining power data, real-time current data, temperature data, and internal resistance data in the battery status data to extract battery status features. The specific method is as follows:

[0016]

[0017] Among them, i represents the i-th time interval, F(i) represents the battery status feature of the i-th time interval, T represents the length of the time interval, t i represents the starting time point of the i-th time interval, R0 represents the nominal internal resistance of the electric heavy truck battery, R(τ) represents the real-time internal resistance of the electric heavy truck battery at the τ-th moment, C(i) represents the remaining power of the electric heavy truck battery at the starting time point of the i-th time interval, C0 represents the nominal power of the electric heavy truck battery, I(τ) represents the real-time current of the electric heavy truck battery at the τ-th moment, ΔT(i) represents the absolute value of the difference in temperature between the termination time point and the starting time point of the electric heavy truck battery in the i-th time interval, and η represents the temperature compensation coefficient of the electric heavy truck battery.

[0018] Furthermore, integrate the real-time speed of the electric heavy truck, variable acceleration driving duration, and variable acceleration driving mileage in the vehicle driving status data to extract vehicle driving features. The specific method is as follows:

[0019]

[0020] Among them, i represents the i-th time interval, D(i) represents the vehicle driving feature of the i-th time interval, T represents the length of the time interval, ti represents the starting time point of the i-th time interval, v(τ) represents the real-time speed of the electric heavy truck at the τ-th moment, m(i) represents the variable acceleration driving mileage of the electric heavy truck in the i-th time interval, h(i) represents the variable acceleration driving time of the electric heavy truck in the i-th time interval, f(i) represents the number of times the electric heavy truck performs variable acceleration driving in the i-th time interval, k v represents the variable acceleration driving speed weight coefficient of the electric heavy truck, k f represents the variable acceleration driving frequency weight coefficient of the electric heavy truck.

[0021] Furthermore, by integrating the real-time load weight and load driving data of the electric heavy truck in the vehicle load status data, the vehicle load characteristics are extracted. The specific method is as follows:

[0022]

[0023] Among them, i represents the i-th time interval, W(i) represents the vehicle load characteristic in the i-th time interval, j represents the j-th load status time period, n(i) represents that there are a total of n(i) load status time periods in the i-th time interval for the electric heavy truck, n(i) is greater than or equal to 1, w(i, j) represents the load weight of the electric heavy truck in the j-th load status time period of the i-th time interval, m(i, j) represents the driving mileage of the electric heavy truck in the j-th load status time period of the i-th time interval, and h(i, j) represents the driving time of the electric heavy truck in the j-th load status time period of the i-th time interval.

[0024] Furthermore, based on the extracted historical electric heavy truck vehicle operation characteristic data, a prediction model for electric heavy truck operation data is constructed to obtain the predicted value of the electric heavy truck vehicle operation characteristic data; the prediction model is an LSTM model, and the specific formula for the update method of the internal state is as follows:

[0025]

[0026] Among them, i represents the i-th time interval, G(i) represents the updated state unit value in the i-th time interval, I(i) represents the output value of the input layer in the i-th time interval, Q(i) represents the output value of the output layer in the i-th time interval, O(i - 1) represents the output value of the LSTM model in the i-th time interval, A(i - 1) represents the input of the LSTM model in the (i - 1)-th time interval, α represents the conventional weight coefficient of the LSTM model, β represents the input weight coefficient of the LSTM model, and λ represents the output weight coefficient of the LSTM model.

[0027] Furthermore, based on the predicted values of the vehicle operation characteristic data of different electric heavy trucks, a charge and discharge demand priority index is set. The specific formula of the charge and discharge demand priority index is as follows:

[0028]

[0029] Among them, i represents the i-th time interval, and YX(i) represents the priority index of the charging and discharging demand of the electric heavy truck in the i-th time interval. represents the predicted value of the battery state characteristic in the i-th time interval. represents the predicted value of the vehicle driving characteristic in the i-th time interval. represents the predicted value of the vehicle load characteristic in the i-th time interval, k F represents the battery state characteristic weight coefficient, k D represents the vehicle driving characteristic weight coefficient, k W represents the vehicle load characteristic weight coefficient.

[0030] Furthermore, based on the priority index of the charging and discharging demand, the charging and discharging power of the electric heavy truck is calculated. The specific calculation method is as follows:

[0031]

[0032] Among them, u represents the u-th electric heavy truck, and P u (i) represents the charging and discharging power of the u-th electric heavy truck in the i-th time interval, m represents the m-th electric heavy truck, X represents that there are a total of X electric heavy trucks that need to perform dynamic distribution of charging and discharging power, and YX u (i) represents the priority index of the charging and discharging demand of the u-th electric heavy truck in the i-th time interval, C u (i) represents the remaining power of the u-th electric heavy truck in the i-th time interval, and Cd(i) represents the remaining capacity of the power grid in the i-th time interval, k cd is the power grid capacity weight coefficient.

[0033] The present invention also proposes a large-scale electric heavy truck charging and discharging collaborative control system, including:

[0034] A vehicle operation data feature extraction module, which is used to collect the vehicle operation data of the electric heavy truck under the operation state and extract the data features of the vehicle operation data of the electric heavy truck.

[0035] A vehicle operation data prediction module, which is used to construct a prediction model of the electric heavy truck operation data according to the extracted historical vehicle operation characteristic data of the electric heavy truck.

[0036] A vehicle charging and discharging collaborative control module, which is used to set the priority index of the charging and discharging demand according to the predicted values of the vehicle operation characteristic data of different electric heavy trucks, and comprehensively determine the charging and discharging collaborative control method of the electric heavy truck.

[0037] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program, when loaded onto the processor, implements any one of the steps of the method for coordinated charging and discharging control of a large-scale electric heavy truck described above.

[0038] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0039] 1. By comprehensively extracting battery state features from battery state data, the present invention can monitor the health status of the battery in real time. By analyzing vehicle driving characteristic data, it can identify the variable acceleration driving conditions during vehicle driving. By extracting vehicle load characteristic data, it can understand the driving conditions and performance of the vehicle under different load conditions, which helps to achieve refined management of the vehicle and improve the operation efficiency of electric heavy trucks;

[0040] 2. By using the LSTM network to establish a prediction model for the operating state of an electric heavy truck, the present invention can more accurately predict the future operating state of the electric heavy truck, including battery state, driving characteristics, and load characteristics. The constructed prediction model can provide decision-making support for the operation, maintenance, and scheduling of electric heavy trucks. By predicting the future operating state of the vehicle, it helps to achieve advance planning of the charging strategy;

[0041] 3. By setting priority indicators for charging and discharging requirements, the present invention can comprehensively consider the battery state, driving characteristics, and load characteristics of an electric heavy truck, thereby more accurately evaluating the charging and discharging requirements of each electric heavy truck. This helps to formulate a more reasonable charging and discharging strategy, avoid blind charging and discharging, improve the charging and discharging efficiency, and dynamically allocate the charging and discharging power according to the priority indicators of the charging and discharging requirements, remaining power, and grid capacity of different electric heavy trucks. This dynamic allocation method can ensure the rational use of grid resources, avoid grid overload or resource waste, and at the same time meet the charging and discharging requirements of electric heavy trucks. Description of the Drawings

[0042] Figure 1 is a step diagram of the method for coordinated charging and discharging control of a large-scale electric heavy truck according to the present invention;

[0043] Figure 2 is a step diagram of the method for dynamic allocation of electric power of the method for coordinated charging and discharging control of a large-scale electric heavy truck according to the present invention.

[0044] Figure 3 is a system module diagram of the system for coordinated charging and discharging control of a large-scale electric heavy truck according to the present invention. Detailed Embodiments

[0045] To make the objectives, technical solutions, and advantages of the present invention clearer, the following will clearly and completely describe the technical solutions of the present invention in conjunction with the accompanying drawings in the embodiments of the present invention. The embodiments described in this application are only a part of the embodiments of the present invention, rather than all embodiments. Based on the spirit of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the protection scope of the present invention.

[0046] Embodiment 1

[0047] As Figure 1 、 Figure 2 shown, a method for coordinated control of charging and discharging of large-scale electric heavy trucks specifically includes the following steps:

[0048] Step 1: Collect vehicle operation data of an electric heavy truck under its operating state, and extract characteristic data of the vehicle operation data of the electric heavy truck.

[0049] Equip the electric heavy truck with in-vehicle sensors in advance; dynamically divide time intervals, where the length of each time interval is T, and the actual size of T is set according to actual needs; collect vehicle operation data in real time through the in-vehicle sensors within each time interval; the vehicle operation data includes battery state data, vehicle driving data, and vehicle load data; upload all the collected vehicle operation data to the cloud data platform through wireless communication technology for storage and analysis; based on the collected battery state data, vehicle driving data, and vehicle load data, extract the corresponding vehicle operation characteristic data, which is used to represent the impacts of the battery state, vehicle driving state, and vehicle load state during the vehicle operation time.

[0050] By dynamically dividing time intervals to collect multiple types of data, the operation information of the electric heavy truck can be comprehensively and accurately obtained. Extracting characteristics from the three aspects of the battery, driving, and load of the electric heavy truck provides a rich and targeted data basis for subsequent analysis, which helps to achieve in-depth understanding and refined management of the operation state of the electric heavy truck.

[0051] Integrate the remaining power data, real-time current data, temperature data, and internal resistance data in the battery state data, and extract the battery state characteristics. The specific formula is as follows:

[0052]

[0053] where i represents the i-th time interval, F(i) represents the battery state characteristic of the i-th time interval, T represents the length of the time interval, and t i$t_i$ represents the starting time point of the $i$-th time interval, $R_0$ represents the nominal internal resistance of the electric heavy truck battery, $R(\tau)$ represents the real-time internal resistance of the electric heavy truck battery at the $\tau$-th moment, $C(i)$ represents the remaining battery charge of the electric heavy truck battery at the starting time point of the $i$-th time interval, denoted as the initial remaining battery charge of the electric heavy truck battery in the $i$-th time interval, $C_0$ represents the nominal battery charge of the electric heavy truck battery, $I(\tau)$ represents the real-time current of the electric heavy truck battery at the $\tau$-th moment, $\Delta T(i)$ represents the absolute value of the temperature difference between the termination time point and the starting time point of the electric heavy truck battery in the $i$-th time interval, denoted as the temperature change of the electric heavy truck battery in the $i$-th time interval, and $\eta$ represents the temperature compensation coefficient of the electric heavy truck battery, which is preset according to the battery material of the electric heavy truck.

[0054] In addition to the method of the present invention, those skilled in the art can also use the following method to extract battery state characteristics:

[0055]

[0056] Among them, $i$ represents the $i$-th time interval, $F(i)$ represents the battery state characteristic of the $i$-th time interval, $C(i)$ is the remaining battery charge of the electric heavy truck battery in the $i$-th time interval, $I(i)$ is the real-time current of the electric heavy truck battery in the $i$-th time interval, $T(i)$ is the real-time temperature of the electric heavy truck battery in the $i$-th time interval, $T_0$ is the nominal battery temperature (the battery temperature at room temperature, usually taken as 25 °C), $R(i)$ is the real-time internal resistance of the electric heavy truck battery in the $i$-th time interval, $R_0$ is the nominal internal resistance of the battery, and $\varepsilon$ is the influence coefficient of temperature on battery characteristics.

[0057] The said method is the second preferred implementation of the technical solution of the present invention. To a certain extent, it can achieve the technical objectives of the present invention, but it does not fully exert the full potential of the present invention's solution. Therefore, it can only achieve the expected effect in certain specific scenarios.

[0058] Define the time when the absolute value of the acceleration of the electric heavy truck during driving is greater than or equal to $a$ as the variable acceleration driving time, the mileage traveled by the electric heavy truck during the variable acceleration driving duration as the variable acceleration driving mileage, and a continuous period of variable acceleration driving time as one variable acceleration driving; extract the vehicle driving characteristics by integrating the real-time speed, variable acceleration driving duration, and variable acceleration driving mileage of the electric heavy truck in the vehicle driving state data. The specific formula is as follows:

[0059]

[0060] Among them, $i$ represents the $i$-th time interval, $D(i)$ represents the vehicle driving characteristic of the $i$-th time interval, $T$ represents the length of the time interval, and $t$ irepresents the starting time point of the i-th time interval, v(τ) represents the real-time speed of the electric heavy truck at the τ-th moment, m(i) represents the variable acceleration driving mileage of the electric heavy truck in the i-th time interval, h(i) represents the variable acceleration driving time of the electric heavy truck in the i-th time interval, f(i) represents the number of times the electric heavy truck performs variable acceleration driving in the i-th time interval, k v represents the variable acceleration driving speed weight coefficient of the electric heavy truck, k f represents the variable acceleration driving frequency weight coefficient of the electric heavy truck.

[0061] In addition to the method of the present invention, those skilled in the art can also use the following methods to extract vehicle driving characteristics:

[0062]

[0063] Among them, i represents the i-th time interval, D(i) represents the vehicle driving characteristics in the i-th time interval, V(i) is the real-time speed of the electric heavy truck in the i-th time interval, V max is the maximum speed of the electric heavy truck, m(i) is the variable acceleration driving mileage of the electric heavy truck in the i-th time interval, m max is the maximum driving mileage of the electric heavy truck throughout the journey, h(i) is the variable acceleration driving duration of the electric heavy truck in the i-th time interval, h max is the maximum variable acceleration driving duration of the electric heavy truck throughout the journey, f(i) is the number of times the electric heavy truck performs variable acceleration driving in the i-th time interval. ζ is the weighted coefficient of the variable acceleration driving frequency, considering the impact of variable acceleration on the power demand of the electric heavy truck.

[0064] The second preferred solution for extracting vehicle driving characteristics proposed by the present invention is an implementation manner of the technical solution of the present invention, but there is a certain gap from the optimal solution in terms of comprehensive performance and applicability.

[0065] Define the time period when the load weight of the electric heavy truck remains unchanged during driving as a load state time period, and extract the vehicle load characteristics by integrating the real-time load weight and load driving data of the electric heavy truck in the vehicle load state data. The specific formula is as follows:

[0066]

[0067] Among them, i represents the i-th time interval, W(i) represents the vehicle load characteristics of the i-th time interval, j represents the j-th load state time period, n(i) represents the total number of load state time periods of electric heavy trucks in the i-th time interval, n(i) is greater than or equal to 1, w(i, j) represents the load weight of the electric heavy truck in the j-th load state time period of the i-th time interval, m(i, j) represents the driving mileage of the electric heavy truck in the j-th load state time period of the i-th time interval, and h(i, j) represents the driving time of the electric heavy truck in the j-th load state time period of the i-th time interval.

[0068] In addition to the method of the present invention, those skilled in the art can also use the following method to extract vehicle load characteristics:

[0069]

[0070] Among them, i represents the i-th time interval, W(i) represents the vehicle load characteristics of the i-th time interval, W max The maximum load capacity of the electric heavy truck, m(i, j) is the driving mileage of the electric heavy truck in the j-th load state time period of the i-th time interval, m max Is the maximum driving mileage of the electric heavy truck throughout the journey, n(i) is the number of load state time periods of the electric heavy truck in the i-th time interval, and γ is the weighted coefficient of the number of load state time periods, reflecting the impact of load on the electric heavy truck.

[0071] The above method is the second preferred implementation for extracting vehicle load characteristics. This method can partially achieve the technical effects of the present invention and solve some technical problems.

[0072] Step 2: Based on the extracted historical operation characteristic data of electric heavy trucks, construct a prediction model for the operation data of electric heavy trucks to obtain the predicted values of the operation characteristic data of electric heavy trucks.

[0073] The vehicle operation characteristic data includes battery state characteristics, vehicle driving characteristics, and vehicle load characteristics. Denote the vehicle operation characteristic data as A(i) = {F(i), D(i), W(i)}. Divide the historical operation characteristic data set of electric heavy trucks into a training set and a control set according to the time line. In this embodiment, the training set accounts for 90% of the historical operation characteristic data set of electric heavy trucks, and the rest is the control set. Normalize the decomposed components in the training set and the control set. Select the operation data characteristics of any historical time interval from the training set as the input data of the prediction model. Based on the LSTM network, establish a prediction model for the operation state of electric heavy trucks. Initialize the weight matrix and bias weight coefficient of the LSTM network. Input the input data into the initialized LSTM network for training. The specific formula for the parameter update method of the LSTM model is as follows:

[0074]

[0075] Among them, i represents the i-th time interval, G(i) represents the value of the state unit after being updated in the i-th time interval, I(i) represents the output value of the input layer in the i-th time interval, Q(i) represents the output value of the output layer in the i-th time interval, O(i - 1) represents the output value of the LSTM model in the i-th time interval, A(i - 1) represents the input of the LSTM model in the (i - 1)-th time interval, α represents the conventional weight coefficient of the LSTM model, β represents the input weight coefficient of the LSTM model, and λ represents the output weight coefficient of the LSTM model.

[0076] Update the parameters of the LSTM model through the backpropagation algorithm, evaluate the prediction performance of the model using the control set, determine the prediction error by calculating the mean absolute error between the LSTM output value and the true value in the corresponding time interval, set the prediction error threshold J, obtain the predicted values for N time intervals using the LSTM model, and when the mean absolute error between the predicted values for these N time intervals and the true values in the corresponding time intervals is less than or equal to J, it is determined that the model training is completed.

[0077] By using the LSTM model to accurately predict the future operating state of the vehicle, plan the charging and discharging strategy in advance, reduce the impact on the power grid, and improve the coordination between the operation of electric heavy trucks and the operation of the power grid.

[0078] Step 3: Based on the predicted values of the vehicle operation characteristic data of different electric heavy trucks, set the priority index for charging and discharging requirements, and calculate the charging and discharging power of the electric heavy truck based on the priority index for charging and discharging requirements.

[0079] Based on the vehicle operation characteristic data of the electric heavy truck in the current time interval, through the trained electric heavy truck operation state prediction model, obtain the predicted value of the vehicle operation characteristic data of the electric heavy truck in the next time interval; comprehensively consider the predicted value of the vehicle operation characteristic data of the electric heavy truck in the next time interval, and set the priority index for charging and discharging requirements. The specific formula for the priority index for charging and discharging requirements is as follows:

[0080]

[0081] Among them, i represents the i-th time interval, YX(i) represents the priority index for charging and discharging requirements of the electric heavy truck in the i-th time interval, represents the predicted value of the battery state characteristic in the i-th time interval, represents the predicted value of the vehicle driving characteristic in the i-th time interval, represents the predicted value of the vehicle load characteristic in the i-th time interval, k F represents the battery state characteristic weight coefficient, k D represents the vehicle driving characteristic weight coefficient, kW It represents the weight coefficient of the vehicle load characteristics. The correlation between each characteristic and the charging and discharging demand (such as the Pearson coefficient) is calculated through historical data, and the weights are allocated according to the correlation ratio, where k F It is set according to the remaining battery power, temperature change, etc. that directly affect the degree of charging and discharging, k D It is set according to the degree of influence of the vehicle driving mode on energy consumption, k W It is set according to the degree of change of the product of load and driving mileage with the battery power, and k F +k D +k W = 1.

[0082] The charging and discharging demand priority index formula synthesizes the predicted values of various characteristics, and can more accurately evaluate the charging and discharging demands of each electric heavy truck. It avoids blind charging and discharging, reasonably arranges the charging and discharging sequence, improves the charging and discharging efficiency, makes full use of the grid resources, and ensures the efficient operation of the electric heavy truck.

[0083] According to the charging and discharging demand priority index, remaining battery power and grid capacity of different electric heavy trucks, the dynamic allocation of the charging and discharging power of electric heavy trucks is determined. The specific calculation method is as follows:

[0084]

[0085] Among them, u represents the u-th electric heavy truck, P u (i) represents the charging and discharging power of the u-th electric heavy truck in the i-th time interval, m represents the m-th electric heavy truck, X represents that there are X electric heavy trucks that need to perform dynamic allocation of charging and discharging power, YX u (i) represents the charging and discharging demand priority index of the u-th electric heavy truck in the i-th time interval, C u (i) represents the remaining battery power of the u-th electric heavy truck in the i-th time interval, Cd(i) represents the remaining capacity of the grid in the i-th time interval, k cd is the grid capacity weight coefficient.

[0086] The charging and discharging power dynamic allocation formula allocates power based on the priority index, remaining battery power and grid capacity. It ensures the reasonable use of grid resources, avoids grid overload, meets the different charging and discharging demands of electric heavy trucks, realizes the two-way optimization of electric heavy trucks and the grid, and improves the overall system stability and reliability.

[0087] Through the above method, the refined dynamic collaborative control of the charging and discharging of electric heavy trucks can be realized.

[0088] Example 2

[0089] Such as Figure 3 shown, a large-scale electric heavy truck charging and discharging collaborative control system specifically includes:

[0090] The vehicle operation data feature extraction module is pre-equipped with on-vehicle sensors for electric heavy trucks, dynamically divides time intervals, and the length of each time interval is T. The actual size of T is set according to actual requirements. In each time window, the vehicle operation data is collected in real time through on-vehicle sensors. The vehicle operation data includes battery status data, vehicle driving data, and vehicle load data. All the collected vehicle operation data is uploaded to the cloud data platform through wireless communication technology for storage and analysis. Based on the collected battery status data, vehicle driving data, and vehicle load data, the corresponding vehicle operation feature data is extracted respectively to represent the impacts of the battery status, vehicle driving status, and vehicle load status during vehicle operation;

[0091] Integrate the remaining power data, real-time current data, temperature data, and internal resistance data in the battery status data to extract battery status features. The specific formula is as follows:

[0092]

[0093] Among them, i represents the i-th time interval, F(i) represents the battery status feature of the i-th time interval, T represents the length of the time interval, t i represents the starting time point of the i-th time interval, R0 represents the nominal internal resistance of the electric heavy truck battery, R(τ) represents the real-time internal resistance of the electric heavy truck battery at the τ-th moment, C(i) represents the remaining power of the electric heavy truck battery at the starting time point of the i-th time interval, denoted as the initial remaining power of the electric heavy truck battery in the i-th time interval, C0 represents the nominal power of the electric heavy truck battery, I(τ) represents the real-time current of the electric heavy truck battery at the τ-th moment, ΔT(i) represents the absolute value of the temperature difference between the termination time point and the starting time point of the electric heavy truck battery in the i-th time interval, denoted as the temperature change of the electric heavy truck battery in the i-th time interval, and η represents the temperature compensation coefficient of the electric heavy truck battery, which is preset according to the battery material of the electric heavy truck.

[0094] Define the time when the absolute value of the acceleration of the electric heavy truck during driving is greater than or equal to a as the variable acceleration driving time, and the mileage traveled by the electric heavy truck during the variable acceleration driving duration is the variable acceleration driving mileage. A continuous period of variable acceleration driving time is recorded as one variable acceleration driving. Integrate the real-time speed, variable acceleration driving duration, and variable acceleration driving mileage of the electric heavy truck in the vehicle driving status data to extract vehicle driving features. The specific formula is as follows:

[0095]

[0096] Among them, i represents the i-th time interval, D(i) represents the vehicle driving feature of the i-th time interval, T represents the length of the time interval, t idenotes the starting time point of the i-th time interval, v(τ) denotes the real-time speed of the electric heavy truck at the τ-th moment, m(i) denotes the variable acceleration driving mileage of the electric heavy truck in the i-th time interval, h(i) denotes the variable acceleration driving time of the electric heavy truck in the i-th time interval, f(i) denotes the number of times the electric heavy truck performs variable acceleration driving in the i-th time interval, k v denotes the variable acceleration driving speed weight coefficient of the electric heavy truck, k f denotes the variable acceleration driving frequency weight coefficient of the electric heavy truck.

[0097] Define the time period when the load weight of the electric heavy truck remains unchanged during driving as a load state time period. Integrate the real-time load weight of the electric heavy truck and the load driving data in the vehicle load state data, and extract the vehicle load characteristics. The specific formula is as follows:

[0098]

[0099] Among them, i represents the i-th time interval, W(i) represents the vehicle load characteristic of the i-th time interval, j represents the j-th load state time period, n(i) represents the total number of load state time periods contained in the electric heavy truck in the i-th time interval, n(i) is greater than or equal to 1, w(i, j) represents the load weight of the electric heavy truck in the j-th load state time period of the i-th time interval, and m(i, j) represents the driving mileage of the electric heavy truck in the j-th load state time period of the i-th time interval, and h(i, j) represents the driving time of the electric heavy truck in the j-th load state time period of the i-th time interval.

[0100] Vehicle operation data prediction module. The vehicle operation characteristic data includes battery state characteristics, vehicle driving characteristics and vehicle load characteristics. Denote the vehicle operation characteristic data as A(i) = {F(i), D(i), W(i)}. Divide the historical electric heavy truck vehicle operation characteristic data set into a training set and a control set according to the time line. In this embodiment, the training set accounts for 90% of the historical electric heavy truck vehicle operation characteristic data set, and the rest is the control set. Normalize the decomposition components in the training set and the control set. Select the vehicle operation data characteristics of any historical time interval in the training set as the input data of the prediction model. Based on the LSTM network, establish a prediction model for the operation state of the electric heavy truck. Initialize the weight matrix and bias weight coefficient of the LSTM network, and input the input data into the initialized LSTM network for training. The specific formula for the update method of the internal state of the LSTM model is as follows:

[0101]

[0102] Among them, i represents the i-th time interval, G(i) represents the value of the state unit after update in the i-th time interval, I(i) represents the output value of the input layer in the i-th time interval, Q(i) represents the output value of the output layer in the i-th time interval, O(i - 1) represents the output value of the LSTM model in the i-th time interval, A(i - 1) represents the input of the LSTM model in the (i - 1)-th time interval, α represents the conventional weight coefficient of the LSTM model, β represents the input weight coefficient of the LSTM model, and λ represents the output weight coefficient of the LSTM model.

[0103] Update the parameters of the LSTM model through the backpropagation algorithm, evaluate the prediction performance of the model using the control set, determine the prediction error by calculating the mean absolute error between the LSTM output value and the true value in the corresponding time interval, set the prediction error threshold J, obtain the predicted values for N time intervals using the LSTM model, and when the mean absolute error between the predicted values for these N time intervals and the true values in the corresponding time intervals is less than or equal to J, it is determined that the model training is completed.

[0104] The vehicle charging and discharging collaborative control module, based on the vehicle operation characteristic data of the electric heavy truck in the current time interval, obtains the predicted values of the vehicle operation characteristic data of the electric heavy truck in the next time interval through the trained electric heavy truck operation state prediction model, and combines the predicted values of the vehicle operation characteristic data of the electric heavy truck in the next time interval to set the charging and discharging demand priority index. The specific formula for the charging and discharging demand priority index is as follows:

[0105]

[0106] Among them, i represents the i-th time interval, and YX(i) represents the charging and discharging demand priority index of the electric heavy truck in the i-th time interval. represents the predicted value of the battery state characteristic in the i-th time interval. represents the predicted value of the vehicle driving characteristic in the i-th time interval. represents the predicted value of the vehicle load characteristic in the i-th time interval, k F represents the battery state characteristic weight coefficient, k D represents the vehicle driving characteristic weight coefficient, k W represents the vehicle load characteristic weight coefficient.

[0107] Determine the dynamic allocation of the charging and discharging power of the electric heavy truck according to the charging and discharging demand priority index, remaining power, and grid capacity of different electric heavy trucks. The specific calculation method is as follows:

[0108]

[0109] Among them, u represents the u-th electric heavy truck, P u$P_{u,i}$ represents the charging and discharging power of the $u$-th electric heavy truck in the $i$-th time interval, $m$ represents the $m$-th electric heavy truck, $X$ represents that there are a total of $X$ electric heavy trucks that need to perform dynamic allocation of charging and discharging power, $Y_X$ u $D_{u,i}$ represents the priority index of the charging and discharging demand of the $u$-th electric heavy truck in the $i$-th time interval, $C$ u $S_{u,i}$ represents the remaining power of the $u$-th electric heavy truck in the $i$-th time interval, $Cd(i)$ represents the remaining capacity of the power grid in the $i$-th time interval, $k$ cd is the power grid capacity weight coefficient.

[0110] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the computer program, when loaded into the processor, implements any one of the steps of the method for coordinated control of charging and discharging of a large-scale electric heavy truck described above.

[0111] As described above, only the preferred specific embodiments of the present invention are provided, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution and inventive concept of the present invention, makes equivalent substitutions or changes, and should be covered by the protection scope of the present invention.

[0112] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the present invention to only the specific embodiments. Obviously, according to the content of this specification, many modifications and changes can be made. This specification selects and specifically describes these embodiments to better explain the principle and practical application of the present invention, so that those skilled in the art in the relevant technical field can understand and utilize the present invention well. The present invention is only limited by the claims and their full scope and equivalents.

Claims

1. A method for coordinated control of charging and discharging of large-scale electric heavy trucks, characterized in that, Including: Step 1: Collect the vehicle operation historical data of the electric heavy truck in the running state, and extract the characteristic data of the vehicle operation historical data of the electric heavy truck; Step 2: Based on the extracted historical vehicle operation characteristic data of the electric heavy truck, construct and train a prediction model for the operation data of the electric heavy truck; Step 3: Collect the real-time vehicle operation data of different electric heavy trucks and extract the characteristic data, and input them into the trained prediction model for the operation data of the electric heavy truck to obtain the predicted values of the vehicle operation characteristic data of different electric heavy trucks; Based on the predicted values of the vehicle operation characteristic data of different electric heavy trucks, set the priority index of the charging and discharging requirements, and calculate the charging and discharging power of the electric heavy truck based on the priority index of the charging and discharging requirements.

2. A method for coordinated control of charging and discharging of large-scale electric heavy trucks according to claim 1, characterized in that: In step 1, collect the vehicle operation historical data of the electric heavy truck in the running state, and extract the characteristic data of the vehicle operation historical data of the electric heavy truck. The specific method is: Dynamically divide the time interval, the length of each time interval is T, and collect the vehicle operation data in real time through on-vehicle sensors within each time interval; the vehicle operation data includes battery state data, vehicle driving data and vehicle load data; Integrate the remaining power data, real-time current data, temperature data and internal resistance data in the battery state data, and extract the battery state characteristics; Integrate the real-time speed of the electric heavy truck, the variable acceleration driving duration and the variable acceleration driving mileage in the vehicle driving state data, and extract the vehicle driving characteristics; Integrate the real-time load weight of the electric heavy truck and the load driving data in the vehicle load state data, and extract the vehicle load characteristics.

3. A method for coordinated control of charging and discharging of large-scale electric heavy trucks according to claim 2, characterized in that: Integrate the remaining power data, real-time current data, temperature data and internal resistance data in the battery state data, and extract the battery state characteristics. The specific method is: Among them, i represents the i-th time interval, F(i) represents the battery state characteristics of the i-th time interval, T represents the length of the time interval, and t i represents the starting time point of the i-th time interval, R0 represents the nominal internal resistance of the electric heavy truck battery, R(τ) represents the real-time internal resistance of the electric heavy truck battery at the τ-th moment, C(i) represents the remaining power of the electric heavy truck battery at the starting time point of the i-th time interval, C0 represents the nominal power of the electric heavy truck battery, I(τ) represents the real-time current of the electric heavy truck battery at the τ-th moment, ΔT(i) represents the absolute value of the temperature difference between the termination time point and the starting time point of the electric heavy truck battery in the i-th time interval, and η represents the temperature compensation coefficient of the electric heavy truck battery.

4. A method for coordinated control of charging and discharging of large-scale electric heavy trucks according to claim 2, characterized in that: Integrate the real-time speed of the electric heavy truck, the variable acceleration driving duration and the variable acceleration driving mileage in the vehicle driving state data, and extract the vehicle driving characteristics. The specific method is: Among them, i represents the i-th time interval, D(i) represents the vehicle driving characteristics of the i-th time interval, T represents the length of the time interval, and t i represents the starting time point of the i-th time interval, v(τ) represents the real-time speed of the electric heavy truck at the τ-th moment, m(i) represents the variable acceleration driving mileage of the electric heavy truck in the i-th time interval, h(i) represents the variable acceleration driving time of the electric heavy truck in the i-th time interval, f(i) represents the number of times the electric heavy truck performs variable acceleration driving in the i-th time interval, k v represents the variable acceleration driving speed weight coefficient of the electric heavy truck, k f represents the variable acceleration driving frequency weight coefficient of the electric heavy truck.

5. A method for coordinated control of charging and discharging of large-scale electric heavy trucks according to claim 2, characterized in that: Integrate the real-time load weight of the electric heavy truck and the load driving data in the vehicle load state data, and extract the vehicle load characteristics. The specific method is: Wherein, i represents the i-th time interval, W(i) represents the vehicle load characteristic of the i-th time interval, j represents the j-th load state time period, n(i) represents that the electric heavy truck in the i-th time interval includes a total of n(i) load state time periods, n(i) is greater than or equal to 1, w(i, j) represents the load weight of the electric heavy truck in the j-th load state time period of the i-th time interval, m(i, j) represents the driving mileage of the electric heavy truck in the j-th load state time period of the i-th time interval, and h(i, j) represents the driving time of the electric heavy truck in the j-th load state time period of the i-th time interval.

6. The method for coordinated control of charging and discharging of a large-scale electric heavy truck according to claim 1, wherein: In step two, a prediction model of the operating data of the electric heavy truck is constructed based on the LSTM network, and the specific formula for updating the model parameters is as follows: Where i represents the i-th time interval, G(i) represents the value of the state unit updated in the i-th time interval, I(i) represents the output value of the input layer in the i-th time interval, Q(i) represents the output value of the output layer in the i-th time interval, O(i - 1) represents the output value of the LSTM model in the i-th time interval, A(i - 1) represents the input of the LSTM model in the (i - 1)-th time interval, α represents the conventional weight coefficient of the LSTM model, β represents the input weight coefficient of the LSTM model, and λ represents the output weight coefficient of the LSTM model.

7. The method for coordinated control of charging and discharging of a large-scale electric heavy truck according to claim 1, wherein: In step three, based on the predicted values of the vehicle operation characteristic data of different electric heavy trucks, a charging and discharging demand priority index is set, and the specific formula of the charging and discharging demand priority index is as follows: Among them, i represents the i-th time interval, and YX(i) represents the priority index of the charging and discharging demand of the electric heavy truck in the i-th time interval. represents the predicted value of the battery state characteristic in the i-th time interval. represents the predicted value of the vehicle driving characteristic in the i-th time interval. represents the predicted value of the vehicle load characteristic in the i-th time interval, k F represents the weight coefficient of the battery state characteristic, k D represents the weight coefficient of the vehicle driving characteristic, k W represents the weight coefficient of the vehicle load characteristic.

8. The method for coordinated control of charging and discharging of a large-scale electric heavy truck according to claim 7, wherein: The charging and discharging power of the electric heavy truck is calculated based on the charging and discharging demand priority index, and the specific calculation method is as follows: Among them, u represents the u-th electric heavy truck, P u (i) represents the charging and discharging power of the u-th electric heavy truck in the i-th time interval, m represents the m-th electric heavy truck, X represents that there are X electric heavy trucks that need to perform dynamic allocation of charging and discharging power, YX u (i) represents the priority index of the charging and discharging demand of the u-th electric heavy truck in the i-th time interval, C u (i) represents the remaining power of the u-th electric heavy truck in the i-th time interval, Cd(i) represents the remaining capacity of the power grid in the i-th time interval, k cd is the power grid capacity weight coefficient.

9. A large-scale electric heavy truck charge-discharge collaborative control system, applied to the large-scale electric heavy truck charge-discharge collaborative control method described in any one of claims 1-8, characterized in that, Including: A vehicle operation data feature extraction module, configured to collect the vehicle operation data of the electric heavy truck in the operating state and extract the data features of the vehicle operation data of the electric heavy truck; A vehicle operation data prediction module, configured to construct a prediction model of the vehicle operation data of the electric heavy truck according to the extracted historical vehicle operation characteristic data of the electric heavy truck; A vehicle charging and discharging coordinated control module, configured to set a charging and discharging demand priority index according to the predicted values of the vehicle operation characteristic data of different electric heavy trucks and comprehensively determine the method for coordinated control of charging and discharging of the electric heavy truck.

10. An electronic device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The computer program, when loaded into the processor, implements the method for coordinated control of charging and discharging of a large-scale electric heavy truck according to any one of claims 1-8.

Images