Hybrid electric vehicle energy management method, device and system and storage medium

Abstract

The invention discloses a hybrid electric vehicle energy management method, device and system and a storage medium. The method comprises the steps that 1, multi-source traffic information and vehicle state information in the vehicle running process are collected; 2, extracting traffic elements influencing the driving of the hybrid electric vehicle according to the multi-source traffic information and the vehicle state information, and constructing the multi-source traffic information of a graph structure; 3, training a vehicle speed prediction model of a long-short term network based on the graph convolutional neural network according to the multi-source traffic information of the graph structure, and predicting the future vehicle speed by using the trained vehicle speed prediction model; and step 4, according to the predicted vehicle speed, performing energy management considering energy consumption and life decline of a vehicle power system. By adopting the technical scheme of the invention, the influence of the traffic elements on the sub-vehicles is fully utilized, the speed sequence which more accords with the reality is predicted, and the energy management economy is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of hybrid electric vehicles, and in particular relates to a hybrid electric vehicle energy management method and device, system, and storage medium. Background Art

[0002] In today's increasingly severe global environmental problems, transportation plays an important role in the production process of human life, and also accounts for a considerable proportion of energy consumption and emissions. Governments of various countries and international environmental protection organizations have focused their attention on the strategic goal of energy conservation and emission reduction, and vigorously advocated green and sustainable transportation. As the core key to breaking through the dilemma of energy and environment, the study of appropriate and efficient energy management strategies for automobiles will promote this process, improve energy efficiency, and reduce energy consumption and emissions.

[0003] With the rapid development of vehicle Internet technology, traffic information can now be obtained in public transportation with the help of the deep integration and application of cutting-edge technologies such as sensors, cloud computing, big data, and communication networks. The ability of all parties such as vehicles, road facilities, and traffic management departments to achieve all-round and real-time information interconnection and interoperability is gradually increasing. It is very important to study vehicle energy management in an information-interconnected networked traffic environment. At present, the energy management method of autonomous hybrid vehicles lacks consideration of the impact of surrounding traffic elements, resulting in unreasonable prediction of future working conditions, which in turn affects the energy consumption economy of the entire vehicle. Summary of the invention

[0004] The technical problem to be solved by the present invention is to provide a hybrid electric vehicle energy management method and device, system and storage medium.

[0005] To achieve the above object, the present invention adopts the following technical solution:

[0006] A hybrid electric vehicle energy management method, comprising:

[0007] Step 1: Collect multi-source traffic information and vehicle status information during vehicle operation;

[0008] Step 2: Extract traffic elements that affect the driving of hybrid electric vehicles based on multi-source traffic information and vehicle status information, and construct multi-source traffic information with a graph structure;

[0009] Step 3: Train a vehicle speed prediction model based on the long-term and short-term network of the graph convolutional neural network according to the multi-source traffic information of the graph structure, and use the trained vehicle speed prediction model to predict the future vehicle speed;

[0010] Step 4: Based on the predicted vehicle speed, energy management is performed taking into account the energy consumption and life degradation of the vehicle power system.

[0011] Preferably, in step S1, the vehicle state and its surrounding traffic state information S are expressed as:

[0012] S=[v,a,θ,d,d1,d2,d front ,d rear ,Type front ,Type rear ,v front ,a front ,t min ,t max ]

[0013] Where: v is the vehicle speed, a is the acceleration, θ is the wheel steering angle, d is the vehicle travel distance, d1 is the distance between the vehicle and the traffic light intersection ahead, d2 is the distance between the vehicle and the nearest bus stop, d front is the distance between the vehicle and the vehicle in front, d rear is the distance between the vehicle and the following vehicle, Type front is the vehicle type of the preceding vehicle, Type rear is the type of the following vehicle, v front is the speed of the front vehicle, a front is the acceleration of the preceding vehicle, t min is the lower limit of time to pass the traffic light intersection, t max The upper limit of the time it takes to pass through a traffic light intersection.

[0014] Preferably, in step S2, the traffic elements affecting the driving of the hybrid vehicle include: the traffic light closest to the front of the target vehicle, bus stop information, the target vehicle and the status of vehicles around the target vehicle, and the vehicle status includes: vehicle speed, acceleration, lateral position and longitudinal position.

[0015] The present invention also provides a hybrid electric vehicle energy management device, comprising:

[0016] An acquisition module is used to collect multi-source traffic information and vehicle status information during vehicle operation;

[0017] An extraction module is used to extract traffic elements that affect the driving of hybrid vehicles based on multi-source traffic information and vehicle status information, and construct multi-source traffic information with a graph structure;

[0018] The prediction module is used to train a vehicle speed prediction model based on a long-term and short-term graph convolutional neural network according to multi-source traffic information in a graph structure, and use the trained vehicle speed prediction model to predict future vehicle speeds;

[0019] The capacity management module is used to perform energy management taking into account the energy consumption and life degradation of the vehicle power system based on the predicted vehicle speed.

[0020] Preferably, the vehicle state and its surrounding traffic state information S are expressed as:

[0021] S=[v,a,θ,d,d1,d2,d front ,d rear ,Type front ,Type rear ,v front ,a front ,t min ,t max ]

[0022] Where: v is the vehicle speed, a is the acceleration, θ is the wheel steering angle, d is the vehicle travel distance, d1 is the distance between the vehicle and the traffic light intersection ahead, d2 is the distance between the vehicle and the nearest bus stop, d front is the distance between the vehicle and the vehicle in front, d rear is the distance between the vehicle and the following vehicle, Type front is the vehicle type of the preceding vehicle, Type rear is the type of the following vehicle, v front is the speed of the front vehicle, a front is the acceleration of the preceding vehicle, t min is the lower limit of time to pass the traffic light intersection, t max The upper limit of the time it takes to pass through a traffic light intersection.

[0023] Preferably, the traffic elements affecting the driving of the hybrid vehicle include: the traffic light closest to the front of the target vehicle, bus stop information, the target vehicle and the status of vehicles around the target vehicle, and the vehicle status includes: vehicle speed, acceleration, lateral position and longitudinal position.

[0024] The present invention also provides a hybrid electric vehicle energy management system, comprising: a memory and a processor, wherein the memory stores a computer program executed by the processor, and the computer program executes the hybrid electric vehicle energy management method when executed by the processor.

[0025] The present invention also provides a storage medium, on which a computer program is stored, and the computer program executes the hybrid electric vehicle energy management method when running.

[0026] The hybrid vehicle energy management of the present invention takes into account the influence of multi-source traffic information and is suitable for a networked traffic environment. It is used to solve the problem that the energy management method of an autonomous driving hybrid vehicle lacks consideration of the influence of surrounding traffic elements, resulting in unreasonable prediction of future operating conditions, which in turn affects the energy consumption economy of the entire vehicle.

[0027] Compared with the prior art, the present invention has the following advantages:

[0028] 1. The present invention extracts the interaction relationship between the target vehicle and other traffic elements in the form of a directed topological graph, analyzes the vehicle operation logic in a complex traffic environment in the form of graph structured data, and adopts the GCN+LSTM network method to effectively extract the temporal and spatial characteristics between the vehicle and other traffic elements when running in the traffic environment, and fits the nonlinear relationship between the vehicle state and multi-source traffic information and the future vehicle speed, thereby improving the rationality of vehicle speed prediction.

[0029] 2. The energy management method of the present invention under the MPC framework integrates the multi-objective optimization functions of vehicle energy consumption, power system life and ride comfort, and uses the vehicle speed provided by the speed prediction model and the DP solver to realize the local optimal energy allocation method, thereby improving the vehicle energy management system in terms of both economy and comfort. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0031] Figure 1 This is a flow chart of a hybrid vehicle energy management method according to an embodiment of the present invention;

[0032] Figure 2 Schematic diagram of the CGN+LSTM network structure. DETAILED DESCRIPTION

[0033] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0034] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0035] Embodiment 1:

[0036] like Figure 1 As shown, an embodiment of the present invention provides a hybrid electric vehicle energy management method, comprising:

[0037] Step 1: Collect multi-source traffic information and vehicle status information during vehicle operation and establish a vehicle driving database:

[0038] The vehicle-mounted sensors and GPS positioning system are used to obtain information such as the speed, acceleration, steering angle and position of the target vehicle. Through vehicle networking technologies such as V2V and V2I, the traffic light status, bus stop location, and surrounding vehicle information are provided. This information data is uploaded to the cloud-based traffic big data platform. Through timestamp alignment, the synchronization of data at the same time point is guaranteed, and finally a vehicle status and traffic environment information database is established to provide a data basis for the subsequent design of the speed prediction model.

[0039] The controlled vehicle state and its surrounding traffic state information S can be expressed as:

[0040] S=[v,a,θ,d,d1,d2,d front ,d rear ,Type front ,Type rear ,v front ,a front ,t min ,t max ]

[0041] Where: v is the vehicle speed, a is the acceleration, θ is the wheel steering angle, d is the vehicle travel distance, d1 is the distance between the vehicle and the traffic light intersection ahead, d2 is the distance between the vehicle and the nearest bus stop, d front is the distance between the vehicle and the vehicle in front, d rear is the distance between the vehicle and the following vehicle, Type front is the vehicle type of the preceding vehicle, Type rear is the type of the following vehicle, v front is the speed of the front vehicle, a front is the acceleration of the preceding vehicle, t min is the lower limit of time to pass the traffic light intersection, t max The upper limit of the time it takes to pass through a traffic light intersection.

[0042] Step 2: Extract traffic elements that affect the driving of hybrid vehicles based on multi-source traffic information and vehicle status information, and construct multi-source traffic information with a graph structure.

[0043] By utilizing the powerful computing power of cloud computing, the massive data uploaded to the cloud traffic big data platform in step 1 can be quickly analyzed and processed, and valuable information and patterns can be extracted in a short time.

[0044] The target vehicle and the surrounding vehicle states of the target vehicle are extracted, including vehicle speed, acceleration, lateral position and longitudinal position, as well as the signal light and bus stop information closest to the target vehicle, including position, signal light phase and timing, to form the original state data set. Among them, the number of vehicles around the target vehicle does not exceed the maximum number of vehicles within a fixed range, that is, only vehicles within a fixed distance before and after the target vehicle are extracted, and the upper limit of the number of vehicles is fixed, and the vehicles farthest from the target vehicle that exceed the upper limit are not considered.

[0045] The number of vehicles selected around the target vehicle should not exceed the maximum number within the fixed range. Then the distance d between the selected front vehicle and the target vehicle is required to be front ∈[0,+L], the distance d between the selected following vehicle and the target vehicle rear ∈[-L,0].

[0046] The original state data set is converted into a directed graph structure data set, with the target vehicle, surrounding vehicles, bus stops, and intersection lights as nodes, and the node features as the state information of the nodes, to establish a node feature data set at each moment.

[0047] The constructed graph structure, that is, the topological structure of the road, can be expressed as:

[0048] G=(V,E)

[0049] Among them, V represents a set of nodes, and E represents a set of "edges" between nodes.

[0050] The feature of each node in the topological structure diagram is used as a vector to represent the state information of the node, and the feature vectors of multiple nodes are represented by matrices. All feature vectors are classified into multiple matrices to form a feature data set, which is used as the input of GCN in the next step. For example, the feature data set of the traffic flow state around the target vehicle node can be expressed as:

[0051]

[0052] Where t_f represents traffic flow; L represents road lane; s k ={d k ,L k ,l k ,v k} represents the state information of any vehicle k outside the controlled vehicle, d k represents the distance from the controlled vehicle and d k ∈[-L,+L],L k Indicates the lane number, l k represents the length of the vehicle and l k ∈[0,l max ];v k represents the speed of the vehicle and vk ∈[0,v max ];s c_v Indicates the status information of the controlled vehicle.

[0053] Step 3: Train the vehicle speed prediction model based on the long-term and short-term network of the graph convolutional neural network according to the multi-source traffic information of the graph structure in step 2, and use the trained vehicle speed prediction model to predict the future vehicle speed sequence

[0054] The speed prediction model is built based on graph convolutional neural network (GCN), such as Figure 2 As shown. Initially, the historical traffic flow information and directed graph information are input into the GCN network to deeply extract the spatial correlation between the traffic flow information on various roads, and to normalize the data set obtained in step 2 to reduce the learning error caused by the difference in data dimensions. Subsequently, the output information from the GCN network is connected with the original traffic information and passed to the long short-term memory (LSTM) network to extract the temporal correlation related to the change of traffic information. Finally, the fully connected layer is used to derive the mapping relationship between the implicit features of the historical traffic flow information and the predicted value.

[0055] Set the propagation rules of GCN:

[0056]

[0057] represents the preprocessing step, To add a self-connected matrix, I N is the identity matrix; is the degree matrix H (l) is the lth layer, θ (l) Contains the parameters of this layer.

[0058] Encode the topological structure of the road network and the attributes on the road to obtain spatial dependencies. For example, select a 2-layer GCN model to obtain spatial characteristics:

[0059]

[0060] Among them, A is the adjacency matrix used to represent the connection between roads; X is the feature matrix; represents the preprocessing step; σ represents the sigmoid function of the nonlinear model; W0∈R P×H represents the weight matrix from input to hidden layer, P is the length of feature matrix, H is the number of hidden units; W1∈R H×T represents the weight matrix from the hidden layer to the output layer; f(X,A)∈R N×TReLU stands for REctified Linear Unit, which is an activation layer frequently used in modern deep neural networks.

[0061] The node feature matrix is ​​used as the input data of the speed prediction model, and the output data is the vehicle speed information. The historical domain and the prediction domain are set, and the historical domain data is divided into a training set and a test set. The GCN-LSTM speed prediction model is trained using the training set, and the parameters in the GCN-LSTM speed prediction model (weights in GCN, parameters in LSTM, etc.) are adjusted. The trained model is applied to the test set, and the performance of the model on new data that has not participated in the training can be intuitively seen.

[0062] By using the trained speed prediction model, combined with the historical and current data of vehicle status and multi-source traffic information, the future vehicle speed data is predicted to obtain the prediction domain, providing a prediction sequence for the lower-level energy management strategy.

[0063] The calculation formulas for each parameter in the LSTM graph are as follows:

[0064] f t =σ f (W f [h t-1 ,x t ]+b f )

[0065] i t =σ i (W i [h t-1 ,x t ]+b i )

[0066] g t =tanh(W g [h t-1 ,x t ]+b g )

[0067] C t =f t C t-1 +i t g t

[0068] o t =tanh(Wo[h t-1 ,x t ]+b g )

[0069] h t =o t tanh(C t)

[0070] Among them, f t 、i t , o t , C t are the output value of the forget gate, the output value of the input gate, the output value of the output gate, and the hidden state respectively; σ is the corresponding gate function; g t is the network iteration; C t The current state is unit; h t is the current output value; h t-1 and C t-1 Respectively represent the output value and state value of the previous time step; x t represents the current input value; W is the parameter of the gate; b is the bias top that needs to be learned during training.

[0071] To improve the prediction performance of the GCN-LSTM speed prediction model, the final output of the network can be used and the true label value y i To calculate the loss, also known as the mean absolute error (MAE), the root mean square error (MSE) is used as the evaluation index of the speed prediction model. n is the number of all predicted values.

[0072]

[0073] The loss function is used to learn and update parameters to minimize the loss function until it converges. The trained model is used to predict the test set and calculate the RMSE. The smaller the RMSE value, the closer the model's predicted value is to the true value, and the better the model's generalization ability is.

[0074] Step 4: Based on the predicted vehicle speed sequence in step 3, energy management considering the energy consumption and life decay of the vehicle power system is implemented.

[0075] Using the predicted vehicle speed sequence in step 3, further analyze the driving conditions that the vehicle may face in the future, such as urban congestion conditions (low speed, frequent starts and stops). According to the vehicle dynamics model and the predicted driving conditions, predict the vehicle's power demand (including drive power, accessory power, etc.) in the future. At the same time, predict the state changes of key components such as the engine and motor (such as the possible range of changes in engine speed, motor torque, etc.) and the change trend of battery SOC.

[0076] The contents of the vehicle longitudinal dynamics model are as follows:

[0077]

[0078] Among them, P needis the required power of the main vehicle, m is the vehicle mass, g is the acceleration of gravity, f is the vehicle rolling resistance coefficient, δ is the vehicle rotation mass conversion coefficient, α is the road slope, ρ is the air density, A is the vehicle frontal area, C D is the air resistance coefficient, u(t) is the vehicle speed at the next moment, and a(t) is the vehicle acceleration at the next moment.

[0079] In order to obtain the instantaneous power of the drive motor and the instantaneous required power of the vehicle, the following model is established:

[0080]

[0081] Where P need (t) is the instantaneous power required by the main vehicle; P m (t) is the instantaneous power actually required by the motor, η m (P need (t)) is the motor efficiency corresponding to the instantaneous output power, and the relevant data can be obtained by looking up the table.

[0082] Establish an MPC-based energy management strategy, generate the future vehicle speed sequence with the speed prediction model trained in step 3, use the vehicle power system energy consumption, life cost and ride comfort as the power allocation target, use the DP algorithm as the solver, and continuously update the vehicle speed prediction sequence and the corresponding prediction model based on the newly acquired actual vehicle speed and other information as time goes by, and achieve the local optimal solution within the rolling optimization range again. Repeat steps 1 to 4 to complete the energy management during vehicle operation.

[0083] The state variable X of the vehicle at time t t and decision variables U t for:

[0084] X t =[v t ,SOC fc,t ,SOC bat,t ,P fc,t ,P bat,t ]

[0085] U t =[P fc,t ,P bat,t ]

[0086] v t Vehicle speed; SOC fc,t State of charge of a fuel cell system; SOC bat,t The state of charge of the vehicle battery; P fc,t The output power of the fuel cell at the time; P bat,t The output power of the vehicle battery at the moment.

[0087] The vehicle speed v predicted at the next moment in the rolling optimization t+1 And the battery power at the next moment is:

[0088]

[0089] Where, m is the vehicle mass; P all,t is the total output power of the vehicle; F d,t is the rolling resistance; F w,t Air resistance; Δt is the time interval; Q fc is a parameter related to the energy reserve of the fuel cell; η fc is the power generation efficiency of the fuel cell; H sup Hydrogen supply rate: H con is the hydrogen consumption rate; P fc,t is the fuel cell output power; P bat,t is the battery output power; η cha , η dis are the battery charging and discharging efficiencies respectively.

[0090] If the DP algorithm takes energy consumption and life cycle cost into consideration, the hydrogen consumption is converted into hydrogen energy cost, the quantitative unit of the objective function is unified, and a multi-objective optimization function is established. The objective function predicted by the model can be expressed as:

[0091] J mpc =minΣ(C all +αC cost +βt+γ(SOC-SOC ref )+δ|I bat |)

[0092] Among them, SOC ref is the reference SOC, α, β, γ and δ are weight coefficients, C all is the equivalent total hydrogen consumption, I bat is the battery current, C cost is the total system cost, and t is the time interval. The control constraints of the DP algorithm are:

[0093] C all =C fc +C bat

[0094] T m,min ≤T m ≤T m,max

[0095] n m,min ≤n m ≤n m,max

[0096] 0 <P fc,min ≤Pfc ≤P fc,max

[0097] -ΔP fc,max ≤ΔP fc ≤ΔP fc,max

[0098] P fc (k+1)=P fc (k)+ΔP fc (k)

[0099] I bat,min ≤I bat ≤I bat,max

[0100] SOC min ≤SOC≤SOC max

[0101] Among them, C all Indicates the equivalent total hydrogen consumption; C fc Represents the energy consumption of the fuel cell system; C bat Indicates battery energy consumption; P fc Represents the output power of the fuel cell system (FCS); T m Indicates the motor torque; n m Indicates the motor speed; I bat Indicates battery current; SOC battery charging state; ΔP fc Indicates the rate of change of output power; the subscripts min and max indicate the upper and lower limits of the physical quantity, respectively.

[0102] Embodiment 2:

[0103] The embodiment of the present invention further provides a hybrid electric vehicle energy management device, comprising:

[0104] An acquisition module is used to collect multi-source traffic information and vehicle status information during vehicle operation;

[0105] An extraction module is used to extract traffic elements that affect the driving of hybrid vehicles based on multi-source traffic information and vehicle status information, and construct multi-source traffic information with a graph structure;

[0106] The prediction module is used to train a vehicle speed prediction model based on a long-term and short-term graph convolutional neural network according to multi-source traffic information in a graph structure, and use the trained vehicle speed prediction model to predict future vehicle speeds;

[0107] The capacity management module is used to perform energy management taking into account the energy consumption and life degradation of the vehicle power system based on the predicted vehicle speed.

[0108] As an implementation of an embodiment of the present invention, the vehicle state and its surrounding traffic state information S are expressed as:

[0109] S=[v,a,θ,d,d1,d2,d front ,d rear ,Type front ,Type rear ,v front ,a front ,t min ,t max ]

[0110] Where: v is the vehicle speed, a is the acceleration, θ is the wheel steering angle, d is the vehicle travel distance, d1 is the distance between the vehicle and the traffic light intersection ahead, d2 is the distance between the vehicle and the nearest bus stop, d front is the distance between the vehicle and the vehicle in front, d rear is the distance between the vehicle and the following vehicle, Type front is the vehicle type of the preceding vehicle, Type rear is the type of the following vehicle, v front is the speed of the front vehicle, a front is the acceleration of the preceding vehicle, t min is the lower limit of time to pass the traffic light intersection, t max The upper limit of the time it takes to pass through a traffic light intersection.

[0111] As an implementation method of an embodiment of the present invention, traffic elements that affect the driving of hybrid vehicles include: the traffic light closest to the front of the target vehicle, bus stop information, the target vehicle and the status of vehicles around the target vehicle, and the vehicle status includes: vehicle speed, acceleration, lateral position and longitudinal position.

[0112] Embodiment 3:

[0113] An embodiment of the present invention further provides a hybrid electric vehicle energy management system, comprising: a memory and a processor, wherein the memory stores a computer program executed by the processor, and the computer program executes a hybrid electric vehicle energy management method when executed by the processor.

[0114] Embodiment 4:

[0115] An embodiment of the present invention further provides a storage medium, on which a computer program is stored, and the computer program executes the hybrid electric vehicle energy management method when running.

[0116] The embodiments described above are only descriptions of the preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should all fall within the protection scope determined by the claims of the present invention.

Claims

1. A hybrid electric vehicle energy management method, characterized in that: include: Step 1: Collect multi-source traffic information and vehicle status information during vehicle operation; Step 2: Extract traffic elements that affect the driving of hybrid electric vehicles based on multi-source traffic information and vehicle status information, and construct multi-source traffic information with a graph structure; Step 3: Train a vehicle speed prediction model based on the long-term and short-term network of the graph convolutional neural network according to the multi-source traffic information of the graph structure, and use the trained vehicle speed prediction model to predict the future vehicle speed; Step 4: Based on the predicted vehicle speed, energy management is performed taking into account the energy consumption and life degradation of the vehicle power system.

2. The hybrid vehicle energy management method according to claim 1, characterized in that: In step S1, the vehicle state and its surrounding traffic state information S are expressed as: S=[v,a,θ,d,d1,d2,d front ,d rear ,Type front ,Type rear ,v front ,a front ,t min ,t max ] Where: v is the vehicle speed, a is the acceleration, θ is the wheel steering angle, d is the vehicle travel distance, d1 is the distance between the vehicle and the traffic light intersection ahead, d2 is the distance between the vehicle and the nearest bus stop, d front is the distance between the vehicle and the vehicle in front, d rear is the distance between the vehicle and the following vehicle, Type front is the vehicle type of the preceding vehicle, Type rear is the type of the following vehicle, v front is the speed of the front vehicle, a front is the acceleration of the preceding vehicle, t min is the lower limit of time to pass the traffic light intersection, t max The upper limit of the time it takes to pass through a traffic light intersection.

3. The hybrid vehicle energy management method according to claim 2, characterized in that: In step S2, the traffic elements that affect the driving of the hybrid vehicle include: the traffic light closest to the front of the target vehicle, bus stop information, the target vehicle and the status of vehicles around the target vehicle, and the vehicle status includes: vehicle speed, acceleration, lateral position and longitudinal position.

4. A hybrid electric vehicle energy management device, characterized in that: include: An acquisition module is used to collect multi-source traffic information and vehicle status information during vehicle operation; An extraction module is used to extract traffic elements that affect the driving of hybrid vehicles based on multi-source traffic information and vehicle status information, and construct multi-source traffic information with a graph structure; The prediction module is used to train a vehicle speed prediction model based on a long-term and short-term graph convolutional neural network according to multi-source traffic information in a graph structure, and use the trained vehicle speed prediction model to predict future vehicle speeds; The capacity management module is used to perform energy management taking into account the energy consumption and life degradation of the vehicle power system based on the predicted vehicle speed.

5. The hybrid vehicle energy management device according to claim 4, characterized in that: The vehicle status and its surrounding traffic status information S are expressed as: S=[v,a,θ,d,d1,d2,d front ,d rear ,Type front ,Type rear ,v front ,a front ,t min ,t max ] Where: v is the vehicle speed, a is the acceleration, θ is the wheel steering angle, d is the vehicle travel distance, d1 is the distance between the vehicle and the traffic light intersection ahead, d2 is the distance between the vehicle and the nearest bus stop, d front is the distance between the vehicle and the vehicle in front, d rear is the distance between the vehicle and the following vehicle, Type front is the vehicle type of the preceding vehicle, Type rear is the type of the following vehicle, v front is the speed of the front vehicle, a front is the acceleration of the preceding vehicle, t min is the lower limit of time to pass the traffic light intersection, t max The upper limit of the time it takes to pass through a traffic light intersection.

6. The hybrid vehicle energy management device according to claim 5, characterized in that: Traffic elements that affect the driving of hybrid vehicles include: the traffic light closest to the front of the target vehicle, bus stop information, the target vehicle and the status of vehicles around the target vehicle. The vehicle status includes: vehicle speed, acceleration, lateral position and longitudinal position.

7. A hybrid vehicle energy management system, characterized in that: include: A memory and a processor, wherein the memory stores a computer program executed by the processor, and when the computer program is executed by the processor, the hybrid vehicle energy management method according to any one of claims 1 to 3 is executed.

8. A storage medium, characterized in that: The storage medium stores a computer program, which executes the hybrid vehicle energy management method according to any one of claims 1 to 3 when running.

Images