Network connection type new energy automobile energy management method

By adopting an energy management method combining high-precision map navigation and real-time traffic information in plug-in hybrid vehicles, and using improved neural networks and controllers to optimize energy distribution, the contradiction between battery life and fuel consumption in the prior art is solved. More efficient energy utilization and cost reduction are achieved.

CN119037385APending Publication Date: 2024-11-29HENAN UNIV OF SCI & TECH
View PDF 0 Cites 2 Cited by

Patent Information

Application Number
CN202411223467.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-09-03
Publication Date
2024-11-29

AI Technical Summary

Technical Problem

The existing energy management methods of plug-in hybrid vehicles and pure electric vehicles are difficult to optimize fuel consumption without damaging battery life, and cannot effectively utilize the power energy of power batteries, and fail to fully consider the battery charging and discharging. cost impact.

Method used

Using vehicle-mounted high-precision map navigation to plan the optimal driving path, combining real-time traffic information at the vehicle-end road end cloud for rapid processing and regular updates, using improved RBF-NN neural network and GMRES-MPC controller, through principal component analysis and clustering Combined and divided driving conditions, optimized the energy management objective function, consider fuel, power consumption and battery decay costs, and achieve reasonable charging and discharge of power batteries.

Benefits of technology

It improves the rationality of energy management and the efficiency of power battery power utilization, extends battery life, and reduces the operating costs of the entire vehicle, ensuring that the energy management strategy is more consistent with the actual driving conditions of the vehicle and traffic information.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119037385A_ABST
    Figure CN119037385A_ABST
Patent Text Reader

Abstract

The invention belongs to the technical field of new energy automobile energy management, and particularly relates to a network connection type new energy automobile energy management method, which can plan an optimal driving path from a departure place to a destination through vehicle-mounted high-precision map navigation, and can fuse real-time traffic information obtained by a vehicle end and a road end cloud end to quickly process, so that the energy management efficiency is improved. The global driving working condition predicted by fusing the historical working condition, the predicted working condition and the standard working condition is more consistent with the actual condition of the road, so that the formulated energy management method is more consistent with the traffic information and the vehicle running state at the current vehicle driving moment; the vehicle energy management method formulated by the management method is more reasonable; and secondly, the reference SOC is obtained through rapid calculation based on driving related parameters and is updated in real time, and the electric energy of the power battery can be better utilized.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of energy management of new energy vehicles, and specifically relates to an energy management method for networked new energy vehicles. Background Art

[0002] Hybrid electric vehicles and pure electric vehicles are both new energy vehicles. Among them, plug-in hybrid electric vehicles (PHEV) have at least two energy sources compared to traditional vehicles or pure electric vehicles, generally including internal combustion engines and electricity stored in batteries or supercapacitors. The biggest difference between plug-in hybrid vehicles and traditional hybrid vehicles is that the battery power is increased and external charging can be achieved when the vehicle is parked. Therefore, plug-in hybrid vehicles need to consume as much power as possible in the power battery during a single driving mission, further reducing fuel consumption on the basis of traditional hybrid systems. However, how to reasonably consume the battery power at the end of the driving mission, thereby optimizing fuel economy, is a problem that needs to be solved.

[0003] For example, the invention patent document with announcement number CN 110696815 B discloses a predictive energy management method for a networked hybrid electric vehicle. According to the method, the target vehicle uploads its own driving condition information to a data processing center through an on-board terminal device; the data processing center plans the optimal driving path of the target vehicle in combination with the collected road information and estimates its complete vehicle speed curve; the target vehicle receives information feedback from the data processing center and sends it to the vehicle controller VCU for optimal energy distribution in combination with its own collected real-time status information; the VCU quickly responds to the received conditions based on the constructed two-layer feedforward neural network model to obtain the corresponding optimal global SOC trajectory; the VCU uses the MPC method to follow the planned SOC trajectory. In this regard, the current energy management method for plug-in hybrid electric vehicles or pure electric vehicles usually uses the road traffic information at a certain moment to predict the future long-term driving conditions of the vehicle, and then determines the reference SOC based on the predicted future long-term driving conditions of the vehicle, and combines the reference SOC and the actual SOC to manage the vehicle's energy. However, since the road traffic information and the vehicle's driving state change in real time, the driving conditions constructed based on the road traffic information at a certain moment may not be consistent with the global actual road traffic information and vehicle driving state, and the impact of battery degradation caused by the charging and discharging of the power battery is not considered in terms of vehicle cost. Most existing PHEV or BEV energy management strategies consider how to minimize fuel consumption without involving battery performance degradation. Since the battery accounts for a large part of the cost of PHEV and BEV, it is an urgent requirement to properly manage the battery charging / discharging to extend its life while still meeting power requirements and economy, so this leads to the vehicle energy management scheme that only takes the economy of the whole vehicle as the optimization target is not reasonable. Summary of the invention

[0004] The purpose of the present invention is to provide a networked new energy vehicle energy management method, which plans the optimal driving route through the on-board high-precision map navigation and integrates the real-time traffic information obtained by the vehicle-side, road-side and cloud-side for rapid processing, and regularly updates it in a short time domain. The formulated vehicle energy management method is more reasonable and can better utilize the electric energy of the power battery; In order to achieve the above object, the present invention adopts the following technical solution: A method for managing energy of a networked new energy vehicle comprises the following steps: S1. First, use the vehicle-mounted high-precision map to input the destination for navigation planning, and obtain the optimal path between the departure point and the destination. When the vehicle is traveling on the path, various location information of the vehicle at the path is obtained based on GPS, ITS vehicle-mounted and road network intelligent devices, including the speed of the traffic flow ahead, the speed limit of the road, the road slope information, the phase of the traffic lights at the intersection, the relative distance to the vehicle ahead, the relative speed to the vehicle ahead, the pedestrian information at the zebra crossing ahead, the traffic control road construction information, etc., to obtain the possible predicted driving conditions of the vehicle; S2. After collecting road data, the driving condition is predicted from the historical driving data of the vehicle, and then the driving condition is divided based on principal component analysis and cluster combination with the standard driving condition to obtain the predicted global driving condition, and further obtain the speed-time curve of the entire journey. In terms of vehicle operating costs, the global economy is optimized while considering the impact of power battery degradation; when the vehicle is actually driving, the predicted global driving condition is updated and reconstructed every set time (such as 200s) according to the information obtained by the Internet of Vehicles, and the entire journey of the vehicle is the entire journey from the departure point to the destination; S3, combining the actual driving speed of the current vehicle and the road condition information ahead, calculate the vehicle demand torque sequence T _req_t+1 ,...T _req_t+Np The improved RBF-NN neural network vehicle speed predictor is used to obtain the future short-term vehicle speed sequence; the predicted short-term speed sequence is input into the power battery reference SOC generator to obtain the reference SOC curve of the global working condition, and the reference SOC curve is updated in real time when the next predicted speed sequence is input; S4. Input the vehicle demand moment torque T_req, the power battery reference SOC and the vehicle current SOC into the GMRES-MPC controller for rolling optimization, and use the improved generalized minimum residual algorithm to solve the energy management target problem, obtain the future short-term vehicle output torque and the power battery output power, and realize the energy management of new energy vehicles; the energy management optimization objective function is: Among them, f(x(t'),u(t'),w(t'))=C f (t')+C e (t')+C b (t') Where E is the energy management optimization objective function, t0 is the current instantaneous simulation sampling time, t' represents a certain moment in the prediction domain, and N p is the prediction horizon step; C f (t') = k f QUR f (t),C e (t') = ke P b (t), Represent the fuel consumption cost, power consumption cost and power battery equivalent degradation cost respectively. If it is a pure electric vehicle, then C f (t') = 0, k f is the fuel price, Q f (t) is the fuel rate, L / s, k e is the electricity price, P b (t) is the battery power P b (t) = VI b -I b 2 R, k b is the purchase price of the power battery, a i is the interpolation coefficient, obtained by interpolation of the battery semi-empirical model experiment, I b is the power battery discharge current.

[0005] Furthermore, in the above method, a neural network speed predictor is also designed, which can use the predicted future short-term speed sequence of the vehicle and the current actual speed of the vehicle for feedback correction; in the above method, a reference SOC generator is also designed, which inputs the reference SOC into the GMRES-MPC controller for real-time optimization feedback.

[0006] Furthermore, the functions of the neural network domain vehicle speed predictor, reference SOC generator and GMRES-MPC controller are integrated into the vehicle controller, which can facilitate rapid response in the time domain.

[0007] Furthermore, the process of constructing the global driving condition includes the following steps: R1. The working conditions of the entire planned route are divided into three types: high speed (v≥60km / h), medium speed (60km / h≤v≤30km / h), and low speed (v≤30km / h). After data collection and transmission, the data is divided according to speed and stored in the short-trip database and the idle database respectively; R2. When constructing the working condition, the principal component analysis and cluster analysis are first performed on the collected working condition fragments, and then these working condition fragments are randomly combined to construct the working condition, so as to avoid the errors caused by the random selection of working condition fragments in the traditional method, and after preprocessing, the calculation time will be greatly reduced.

[0008] Furthermore, a short trip can be defined as the driving part from the beginning to the end of the vehicle's operation, and idling can be defined as the driving part in which the vehicle's engine is in operation but the vehicle speed is zero; the cumulative frequency method proposed by WLTP is used to establish a cumulative frequency graph through the fragment database, and then short trip fragments and idle fragments are selected; then, a spare short trip fragment database for each type of operating condition is determined, and the short trip fragments therein are randomly combined into spare typical driving conditions, the Euclidean distance of the speed-acceleration distribution is calculated, and the spare typical driving condition with the smallest Euclidean value is selected as the optimal driving condition of this type.

[0009] Furthermore, in the above method, combined with the vehicle speed-time relationship curve of the entire journey of the vehicle, the objective function of the power battery SOC is solved by using the improved minimum residual optimization (GMRES) algorithm under constraints. The objective function of the power battery SOC is: l(x(t0+N p -1),u(t0+N p -1),w(t0+N p -1))=α(SOC(t0+N p -1)-SOC r (t0+N p -1)) 2 Where α is the weight coefficient used to minimize the error between the actual vehicle SOC and the reference SOC to achieve local optimal control, SOC(t0+N p -1) is the actual SOC of the vehicle at that moment, SOC r (t0+N p -1) is the reference SOC of the vehicle at that moment; The constraints are: Where SOC is the SOC of the power battery, min represents the lower limit, and max represents the upper limit; P eng , P ele are engine power and motor power respectively; T eng , T ele , T req , T bra They are engine torque, motor torque, vehicle demand torque and mechanical brake torque respectively; ω eng ,ω ele They are engine speed and motor speed respectively; the vehicle demand torque is obtained according to the speed-time relationship curve of the vehicle's entire journey.

[0010] The beneficial effects of the present invention are as follows: compared with the existing conventional automobile energy management methods, the networked new energy automobile energy management method provided by the present invention can plan the optimal driving route from the departure point to the destination through the on-board high-precision map navigation, and integrate the real-time traffic information obtained from the vehicle-side, road-side and cloud-side for rapid processing, and update it regularly in a short time domain. At the same time, the global driving conditions predicted by integrating historical conditions, predicted conditions and standard conditions are more consistent with the actual road conditions, further ensuring that the formulated energy management method is more consistent with the traffic information and vehicle operating status at the current vehicle driving moment. At the same time, considering the cost impact brought by battery degradation, the vehicle energy management method formulated by the management method of the present invention is more reasonable; secondly, the reference SOC is quickly calculated based on driving-related parameters and updated in real time, which can better utilize the power of the power battery. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 is a schematic diagram of the energy management method for networked new energy vehicles in an embodiment of the present invention; Figure 2 is a flow chart of offline training and online prediction of vehicle speed by a neural network vehicle speed predictor in an embodiment of the present invention; Figure 3 Schematic diagram of the topological network structure of the improved RBF-NN neural network prediction model in an embodiment of the present invention; Figure 4 is a schematic diagram of the rolling optimization principle of the GMRES-MPC controller in an embodiment of the present invention; Figure 5 It is a simplified schematic diagram of energy management of the networked new energy vehicle energy management method in an embodiment of the present invention. DETAILED DESCRIPTION

[0012] Specific embodiment 1: The technical scheme of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of the present invention. It should be noted that: in the present invention, if there is no special description, all the embodiments and preferred implementation methods mentioned herein can be combined with each other to form a new technical scheme. In the present invention, if there is no special description, all the technical features and preferred features mentioned herein can be combined with each other to form a new technical scheme. The "scope" disclosed in the present invention is in the form of lower limit and upper limit, which can be one or more lower limits, and one or more upper limits, respectively. Preferably, the reaction method herein is carried out sequentially. Unless otherwise specified, the professional and scientific terms used herein have the same meaning as those familiar to those skilled in the art.

[0013] The energy management method for networked new energy vehicles provided in this embodiment is as shown in the attached specification. Figure 1 As shown, the method of this embodiment is described in detail below by taking the application of the method of this embodiment to a plug-in hybrid electric vehicle as an example. Of course, the method of this embodiment can also be applied to ordinary hybrid electric vehicles.

[0014] As the instruction manual Figure 1 As shown, a networked new energy vehicle energy management method of this embodiment includes the following steps: Step 1: Use the vehicle-mounted high-precision map to input the destination for navigation planning, and obtain the optimal route between the departure point and the destination. When the vehicle is traveling on the route, various location information and vehicle status information of the vehicle at the location on the route are obtained based on GPS, ITS vehicle-mounted and road network intelligent devices; As other implementation methods, for example, networked hybrid buses, since the bus routes are fixed, the departure and destination are known, and it can be more practically applied to bus models, and the buses can communicate with road-side equipment in real time, so the high-precision map navigation part is replaced, and the remaining steps are as follows; wherein, the location information includes the front traffic speed, road speed limit, road slope information, intersection traffic light phase, relative distance to the front vehicle, relative speed to the front vehicle, zebra crossing pedestrian information ahead, traffic control road construction information, etc. obtained by the on-board laser radar, millimeter wave radar, camera and roadside ITS equipment, so as to achieve the maximum amount of traffic information intake; adopt improved fusion rules The DS evidence theory is used to fuse the above traffic information. Its basic principle is: using multiple sensors as data sources, combining the acquired information in pairs instead of fusing all the information at once, assigning local conflicts to corresponding propositions according to the most recent fusion results, instead of ignoring the information hidden in the conflicting evidence. When facing conflicting evidence information, the evidence can be preprocessed through the relationship between the evidence, so as to achieve the purpose of reducing the error rate of data fusion. Finally, according to the basic probability distribution function, the reasonable judgments are fused together through the corresponding combination rules. Since the above acquired data is in the spatial domain, the spatial domain must be transformed into the time domain before it can be applied to the actual vehicle. Step 2: Combine the historical driving data of the vehicle and the predicted driving conditions with the standard driving conditions to perform combined division based on principal component analysis and clustering, predict the global driving conditions, and predict the vehicle speed-time curve of the entire journey; and use the information flow obtained by the Internet of Vehicles to update and reconstruct at set intervals (such as 200s), so that the entire journey of the vehicle is the optimal planned route from the departure point to the destination; the division and reorganization of the working conditions are as follows: the vehicle data transmitted by the real-time on-board intelligent information terminal is pre-processed by distortion elimination, lost data reconstruction, and validity judgment; the working conditions of the planned route are divided into three types of working conditions: high speed, medium speed, and low speed. After data collection and transmission, the data is divided according to speed and stored in the short-trip database and the idle database respectively; principal component analysis and cluster analysis are performed on the collected working condition fragments; among them, principal component analysis adopts the characteristic parameter method, for example, in this study, a combination of 20 characteristic parameters is selected to describe the characteristics of the short-trip fragment; using all these parameters will inevitably cause large calculation amount and time-consuming calculation problems; the characteristic parameters are reduced in dimension by the principal component analysis method to reduce the number of variables. Taking high-speed working conditions as an example, 18 characteristic parameters were selected in the MATLAB program to perform principal component analysis on the short-trip segments with a duration of 14 seconds in the spare segment database. For example, the cumulative contribution rate of the first five principal components exceeded 90%, which can meet the development needs of the actual vehicle working conditions. Table 1 shows the characteristic parameters with cumulative contribution rate exceeding 90%. Serial number Characteristic parameters meaning 1 <![CDATA[P acc ]]> Acceleration ratio 2 <![CDATA[P mod ]]> Reduction ratio 3 <![CDATA[V max ]]> Maximum speed 4 <![CDATA[P 0-25 ]]> Low speed area ratio 5 <![CDATA[P 25-50 ]]> Medium speed area ratio Furthermore, based on the principal component analysis results of these operating condition fragments, the K-means clustering algorithm is used to cluster the samples, and the K-means function in MATLAB is used to implement the clustering operation. M fragments that are close to the cluster center are selected as candidate fragments for optimization. The fragments of different time lengths in the short-trip fragment database are clustered and analyzed, and then the operating condition fragments are recombined into a spare operating condition fragment database according to the clustering results; the optimization operation after random combination of the spare database is performed to find the operating condition with the closest Euclidean distance to the speed-acceleration distribution of the overall sample, and obtain the typical global driving condition of the entire planned driving route; finally, considering the limitations of information transmission in practical applications, it is assumed that the obtained traffic information is updated every 200 seconds. According to the real-time updated traffic data, the global driving condition is reconstructed every 200 seconds, and the reconstruction process is the same as the previous two steps; after the update is completed, the operating condition combined last time is deleted to avoid conflicts; As another implementation method, the amount of traffic information obtained can be adjusted according to actual conditions, the classification of the operating conditions used in the combined operating conditions can also be reclassified according to actual conditions, and the selection of characteristic parameters can also be adjusted according to the actual vehicle model; Step 3: Calculate the vehicle's required torque sequence based on the actual driving speed of the current vehicle and the road conditions ahead. The improved RBF-NN neural network speed predictor is used to obtain the future short-term driving speed sequence of vehicles. As the instruction manual Figure 2 As shown, the Figure 2 This is a flowchart of offline training and online prediction of vehicle speed by a neural network speed predictor; wherein, the actual vehicle speed sequence at the current moment is processed by online calling the trained offline neural network model, and the vehicle speed sequence in the short-term domain in the future is predicted; wherein the mode of speed prediction is: In the formula, Represents the nonlinear mapping relationship in the neural network, v' t+j Represents the predicted speed at the jth moment in the future; In this embodiment, the improved RBF-NN neural network model adopts a three-layer network structure, including an input layer, a hidden layer and an output layer. Since the Gaussian function has outstanding nonlinear approximation ability, it is set as the basis function of the hidden layer, and linear functions are used in the input layer and the output layer. After a certain number of offline learning and training, the RBF-NN neural network model can accurately predict the future short-time domain speed sequence. Among them, the training data for training the RBF-NN neural network model is the historical operating data of the vehicle; As the instruction manual Figure 3As shown, the attached Figure 3 It is the topological structure diagram of the improved RBF-NN neural network; As other implementation methods, RBF neural network models with other network structures can be selected according to actual needs; Step 4. In order to facilitate application in actual applications, the SOC reference in the predicted sequence can be calculated using the SOC reference generator by predicting the speed sequence. Only the vehicle's initial SOC, terminal SOC and target driving distance are required, which are easy to obtain before the start of the journey, so it is practical in actual applications. According to the predicted short-time domain speed sequence, it is input into the power battery reference SOC generator, and the reference SOC curve is updated in real time when the next predicted speed sequence is input, and finally the reference SOC of the global working condition is obtained; SOC0 is the initial SOC before the vehicle departs, SOC e is the vehicle end-of-trip SOC, D all is the total distance between the departure point and the destination of the optimal planning route of high-precision map navigation, D(t0) is the real-time cumulative driving distance recorded up to sampling time t0 during driving, and SOC r (t0+j) is the SOC reference value in the jth prediction, j=0,1,2...,N p , Δt is the same sample time step as the speed predictor; Step 5: Input the required torque of the vehicle at time t', the reference SOC of the power battery, and the current SOC of the vehicle into the GMRES-MPC controller for rolling optimization. The energy management target problem is solved using the improved generalized minimum residual algorithm to obtain the future short-term vehicle output torque and the output power of the power battery. The objective function of the power battery SOC is constructed with the lowest operating cost of the entire vehicle trip and the impact of battery degradation considered. The objective function of the power battery SOC is: P b (t) = VI b -I b 2 R where J is the objective function of the power battery SOC, f(x(t'),u(t'),w(t')) is the instantaneous objective function, x(t'),u(t'),w(t') are the engine output torque, power battery SOC, and power battery equivalent decay power, respectively. If it is a pure electric vehicle, the x(t') part does not need to be solved; l(x(t0+N p -1),u(t0+N p -1),w(t0+N p-1)) is the instantaneous terminal function, t0 is the current instantaneous simulation sampling time, t' represents a certain moment in the prediction domain, N p is the prediction horizon step; k f is the fuel price, Q f (t) is the fuel rate, L / s, k e is the electricity price, P b (t) is the battery power, k b is the purchase price of the power battery, a i is the interpolation coefficient, obtained by interpolation of the battery semi-empirical model experiment, I b is the power battery discharge current; α is the weight coefficient used to minimize the error between the actual vehicle SOC and the reference SOC to achieve local optimal control; SOC(t0+N p -1) is the actual SOC of the vehicle at that moment, SOC r (t0+N p -1) is the reference SOC of the vehicle at that moment; among them, the functions of the short-term vehicle speed predictor, reference SOC generator and GMRES-MPC controller are all integrated into the vehicle controller, which can conveniently use its high computing power for fast response in the time domain; as shown in the attached manual Figure 4 As shown, the attached Figure 4 It is the schematic diagram of the rolling optimization of the GMRES-MPC controller; the constraint condition of the objective function J of the power battery SOC is: Where SOC is the SOC of the power battery, min represents the lower limit, and max represents the upper limit; P eng , P ele are engine power and motor power respectively; T eng , T ele , T req , T bra They are engine torque, motor torque, vehicle demand torque and mechanical brake torque respectively; ω eng ,ω ele are the engine speed and motor speed respectively; if it is a pure electric vehicle, the constraints are: The vehicle demand torque is obtained based on the vehicle speed-time relationship curve of the entire vehicle journey; In this embodiment, the improved generalized minimum residual algorithm is used to solve the objective function of the power battery SOC. The control variable is the engine torque T e , the state variable is the power battery SOC. If it is a pure electric vehicle, only the state variable part needs to be solved. First, the objective function J of the power battery SOC is solved under the constraint conditions to obtain the output power P of the power battery b (t); In this embodiment, the improved generalized minimum residual algorithm is different from the traditional generalized minimum residual algorithm and the traditional iterative optimization algorithm. The algorithm avoids calculating the Jacobian matrix and the Hessian matrix, thereby showing extremely high computational efficiency, and introduces external penalty constraints to solve the problem that the traditional algorithm cannot directly handle nonlinear inequality constraints and thus cannot be applied to energy management; Step 6. Combine the actual SOC of the power battery, the reference trajectory of the power battery SOC of the vehicle throughout the journey, and the future short-term speed sequence of the vehicle to predict the future output torque and power battery output power of the vehicle, so as to realize the energy management of new energy vehicles; and combine the actual speed of the vehicle to feedback optimize the RBF-NN neural network model, so as to feedback correct the future short-term speed sequence of the vehicle, so as to ensure that the predicted result in the next time domain is more in line with the actual situation, so that the predicted result is closer to the actual vehicle driving.

[0015] In this embodiment, the energy management optimization objective function is solved by using an improved generalized minimum residual algorithm to obtain the vehicle output torque and the output power of the power battery in the future short-term domain, thereby realizing energy management of new energy vehicles; wherein, the energy management optimization objective function is: Among them, f(x(t'),u(t'),w(t'))=C f (t')+C e (t')+C b (t') Where E is the energy management optimization objective function, t0 is the current instantaneous simulation sampling time, t' represents a certain moment in the prediction domain, and N p is the prediction horizon step; C f (t') = k f QUR f (t),C e (t') = k e P b (t), Represent the fuel consumption cost, power consumption cost and power battery equivalent degradation cost respectively. If it is a pure electric vehicle, then C f (t') = 0, k f is the fuel price, Q f (t) is the fuel rate, L / s, k e is the electricity price, P b (t) is the battery power P b (t) = VI b -I b 2 R, k b is the purchase price of the power battery, ai is the interpolation coefficient, obtained by interpolation of the battery semi-empirical model experiment, I b is the power battery discharge current; As the instruction manual Figure 1 As shown, the output torque and power battery output power of the vehicle at the future demand time predicted in step 6 are input into the new energy vehicle model (i.e., PHEV model or BEV model) to realize energy management of the new energy vehicle.

[0016] To sum up, the combined operating condition of the present embodiment is formulated based on the multi-source traffic information acquired by the on-board high-precision map navigation and real-time sensor collection, combined with historical driving condition data and standard operating conditions, and can be updated and corrected according to real-time traffic information, thereby ensuring that the constructed global driving condition is consistent with the actual road conditions; the rolling optimization function of the present embodiment further ensures that the formulated vehicle energy management method is consistent with the traffic information of the vehicle's current driving section and the actual operating status of the vehicle, and the formulated vehicle energy management plan is more reasonable; secondly, the reference trajectory of the power battery SOC of the vehicle's entire journey is also formulated based on the complete journey of the car, which can better utilize electrical energy and avoid multiple charging and discharging of the power battery, thereby extending the service life of the power battery; and the impact of battery degradation is taken into account in the use cost of the entire vehicle, which is more in line with actual use needs. The above is only a preferred embodiment of the present invention and does not limit the present invention in any form. Although the present invention has been disclosed as a preferred embodiment as above, it is not used to limit the present invention. Any technician familiar with this profession can make some changes or modify the technical contents disclosed above into equivalent embodiments without departing from the scope of the technical solution of the present invention. However, any simple modification, equivalent change and modification made to the above embodiments according to the technical essence of the present invention without departing from the content of the technical solution of the present invention still fall within the scope of the technical solution of the present invention.

Claims

1. A networked new energy vehicle energy management method, It is characterized in that The following steps are involved: S1. First, use the vehicle-mounted high-precision map to input the destination for navigation planning, and obtain the optimal path between the departure point and the destination. When the vehicle is traveling on the path, various location information of the vehicle on the path is obtained based on GPS, ITS vehicle-mounted and road network intelligent devices, including the speed of the traffic flow ahead, the road speed limit, the road slope information, the phase of the traffic lights at the intersection, the relative distance to the vehicle ahead, the relative speed to the vehicle ahead, the pedestrian information at the zebra crossing ahead, and the traffic control road construction information, to obtain the possible predicted driving conditions of the vehicle; S2. After collecting road data, the driving condition is predicted from the historical driving data of the vehicle, and then the predicted global driving condition is obtained based on principal component analysis and cluster combination division with the standard driving condition, and the speed-time curve of the entire journey is further obtained. In terms of vehicle operating costs, the global economy is optimized while considering the impact of power battery degradation; when the vehicle is actually driving, the predicted global driving condition is updated and reconstructed at set intervals based on the information obtained from the Internet of Vehicles, and the entire journey of the vehicle is the entire journey from the departure point to the destination; S3, combining the actual driving speed of the current vehicle and the road condition information ahead, calculate the vehicle demand torque sequence The improved RBF-NN neural network vehicle speed predictor is used to obtain the future short-term vehicle speed sequence; the predicted short-term speed sequence is input into the power battery reference SOC generator to obtain the reference SOC curve of the global working condition, and the reference SOC curve is updated in real time when the next predicted speed sequence is input; S4. Input the vehicle demand moment torque T_req, the power battery reference SOC and the vehicle current SOC into the GMRES-MPC controller for rolling optimization, and use the improved generalized minimum residual algorithm to solve the energy management target problem, obtain the future short-term vehicle output torque and the power battery output power, and realize the energy management of new energy vehicles; the energy management optimization objective function is: where f(x(t'), u(t'), w(t')) = C f (t') + C e (t') + C b (t') Where E is the energy management optimization objective function, t 0 is the current instantaneous simulation sampling time, t' represents a certain moment in the prediction domain, N p is the prediction horizon step; Represent the fuel consumption cost, power consumption cost and power battery equivalent degradation cost respectively. If it is a pure electric vehicle, then C f (t') = 0, k f is the fuel price, Q f (t) is the fuel rate, L / s, k e is the electricity price, P b (t) is the battery power P b (t) = VI b -I b 2 R, k b is the purchase price of the power battery, a i is the interpolation coefficient, obtained by interpolation of the battery semi-empirical model experiment, I b is the power battery discharge current.

2. A networked new energy vehicle energy management method according to claim 1, It is characterized in that In the above method, a neural network speed predictor is also designed, which can use the predicted future short-term speed sequence of the vehicle and the current actual speed of the vehicle for feedback correction; at the same time, a reference SOC generator is also designed, which inputs the reference SOC into the GMRES-MPC controller for real-time optimization feedback.

3. A networked new energy vehicle energy management method according to claim 2, It is characterized in that The functions of the neural network domain vehicle speed predictor, reference SOC generator and GMRES-MPC controller are integrated into the vehicle controller to facilitate rapid response in the time domain.

4. A networked new energy vehicle energy management method according to claim 1, It is characterized in that The process of constructing the global driving condition includes the following steps: R1. The working conditions of the entire planned route are divided into three types: high speed (v≥60km / h), medium speed (60km / h≤v≤30km / h), and low speed (v≤30km / h). After data collection and transmission, the data is divided according to speed and stored in the short-trip database and the idle database respectively; R2. When constructing the working condition by combination, the principal component analysis and cluster analysis are first performed on the collected working condition fragments, and then these working condition fragments are randomly combined to construct the working condition, so as to avoid the errors caused by the random selection of working condition fragments in the traditional method, and after preprocessing, the calculation time will be reduced.

5. A networked new energy vehicle energy management method according to claim 1, It is characterized in that A short trip can be defined as the portion of a vehicle's travel from the start to the end of its operation, and idling can be defined as the portion of a vehicle's travel when the engine is in operation but the vehicle speed is zero; The cumulative frequency method proposed by WLTP is used to establish a cumulative frequency graph through the fragment database, and then short-trip fragments and idling fragments are selected; then the spare short-trip fragment database for each type of operating condition is determined, and the short-trip fragments are randomly combined into spare typical driving conditions, and the Euclidean distance of the speed-acceleration distribution is calculated, and the spare typical driving condition with the smallest Euclidean value is selected as the optimal driving condition of this type.

6. A networked new energy vehicle energy management method according to claim 1, It is characterized in that In the above method, combined with the vehicle speed-time relationship curve of the entire vehicle journey, the improved minimum residual optimization (GMRES) algorithm is used under constraints to solve the objective function of the power battery SOC, and the objective function of the power battery SOC is: l(x(t 0 +N p -1),u(t 0 +N p -1),w(t 0 +N p -1))=α(SOC(t 0 +N p -1)-SOC r (t 0 +N p -1)) 2 Where α is the weight coefficient used to minimize the error between the actual vehicle SOC and the reference SOC to achieve local optimal control. 0 +N p -1) is the actual SOC of the vehicle at that moment, SOC r (t 0 +N p -1) is the reference SOC of the vehicle at that moment; The constraints are: Where SOC is the SOC of the power battery, min represents the lower limit, and max represents the upper limit; P eng , P ele They are engine power and motor power respectively; T eng , T ele , T req , T bra They are engine torque, motor torque, vehicle demand torque and mechanical brake torque respectively; ω eng ,ω ele They are engine speed and motor speed respectively; The vehicle demand torque is obtained based on the vehicle speed-time relationship curve of the entire vehicle journey.

Citation Information

Cited By

  • Energy management method for vehicle

    CN120056959A

  • Vehicle energy management system and method based on dynamic path feedback and multimode optimization

    CN121224660A