Power system energy management method and equipment based on hydrogen energy hybrid power

Abstract

The invention discloses a power system energy management method and equipment based on hydrogen energy hybrid power, belongs to the technical field of energy management, and is used for solving the technical problems that the random load characteristic of field operation of an existing tractor is complex, and a hybrid power tractor is uneven in ploughing and sowing under the complex terrain operation working condition and the tillage state. The method comprises the following steps: collecting multi-modal data of a farmland environment; performing vehicle speed prediction processing on the tractor under related environment characterization to obtain a dynamic prediction vehicle speed of the tractor; carrying out rolling optimization processing on the dynamic prediction vehicle speed under related output power to obtain the driving demand power of the tractor; and superposing the driving required power and the pre-obtained tillage tool required power to obtain the total required power of the power system, decomposing the required power in a frequency domain according to a formulated variational mode decomposition rule, and endowing the required power to a driving power generation system formed by a fuel cell, a power cell, a hydrogen internal combustion engine and a generator.

Description

Technical Field

[0001] The present application relates to the field of power reconstruction, and in particular to a method and device for energy management of a power system based on hydrogen hybrid power. Background Art

[0002] The National Agricultural Machinery Equipment Innovation Center has successfully developed China's first 5G+ hydrogen fuel electric tractor. Compared with lithium batteries, hydrogen fuel cells have higher energy density and longer endurance, but the response of a single fuel cell is slow in the high-load and high-humidity farmland environment, affecting efficiency.

[0003] It can be seen that pure electric tractors have problems such as short cruising range and high cost. Hydrogen fuel cell tractors have the disadvantages of high cost and performance that cannot meet the variable load changes in the field. Hybrid tractors have poor energy-saving effects. Therefore, it is necessary to develop a new power system to meet the needs of field operations.

[0004] Some research institutions have made significant progress in energy management strategies for electric tractors, particularly in dynamic energy scheduling and real-time control strategies. Nanjing Agricultural University (CN202411006230.9) designed an optimal power allocation sequence based on minimizing equivalent fuel consumption. This optimal power allocation sequence is used to allocate power to the tractor's energy sources. The power allocation model is constructed based on constraints, an optimization objective function, and an optimal equivalent factor. This optimal equivalent factor is a global optimal solution obtained through optimization using a monkey swarm algorithm. Determining the optimal equivalent factor is crucial to the accuracy of the model, but its calculation may rely on extensive experimental data or empirical values. Patent: A Composite Power Source Energy Management Method Based on BP Neural Network Speed ​​Prediction (Method 201910051538.8). This method collects and processes current vehicle driving status and historical speed information as input parameters for a neural network model to obtain a predicted speed for the future time period. Based on this predicted speed information and combined with the vehicle model, the motor power demand is calculated. A model predictive control algorithm is then used to determine the optimal control variable within the predicted time domain. However, this method uses only a simple neural network and does not consider the impact of the environment on the system. In 2020, the National Agricultural Machinery Equipment Innovation Center successfully developed China's first 5G+ hydrogen fuel electric tractor, applying intelligent vehicle control technology to the tractor, which can optimize the power supply system according to different loads and operating conditions. When the load is light, only the fuel cell provides power, and when the load is heavy, the fuel cell and lithium battery work together to provide power. However, under complex working conditions (such as frequent load fluctuations), the real-time and accuracy of the energy management strategy still need to be improved.

[0005] Tractor field operations are characterized by complex random load characteristics and large load fluctuations. Hybrid tractors face insufficient real-time and precision in their energy management systems under these complex operating conditions. During heavy-load operations, the high power from multiple energy sources is unevenly distributed, resulting in reduced energy efficiency. Intelligent driving devices (such as sensors, controllers, and autonomous driving systems) can monitor the tractor's operating conditions and environmental information in real time, providing data support for the energy management system. However, current energy management strategies are not yet tightly integrated with intelligent driving devices. Key research areas for hydrogen hybrid tractors include coordinating energy distribution among multiple energy sources, integrating intelligent driving devices, and developing appropriate energy management strategies to achieve efficient and intelligent energy distribution under complex operating and environmental conditions. Summary of the Invention

[0006] The present invention provides a hydrogen hybrid powertrain energy management method and device to address the following technical issues: Existing tractors experience complex random load characteristics and large load fluctuations during field operations. Hybrid tractors, under complex operating conditions, lack real-time and precision in their energy management systems. During heavy-load operations, high power from multiple energy sources is unevenly distributed, resulting in reduced energy efficiency. Field terrain is complex, and large speed fluctuations due to the terrain can easily lead to uneven plowing and seeding. Maintaining constant speed and efficient power distribution are urgent challenges. Based on these challenges, a hydrogen hybrid tractor and its energy management strategy were designed.

[0007] The embodiments of this application adopt the following technical solutions:

[0008] On the one hand, an embodiment of the present application provides a power system energy management method based on hydrogen hybrid power, including: collecting multimodal data of the farmland environment, and constructing a data-driven farmland environment cognition model based on key features of the farmland environment; according to the dynamic environmental features in the farmland environment cognition model, and based on the TCN time series speed prediction model, the tractor is subjected to vehicle speed prediction processing under relevant environmental characterization to obtain the dynamic predicted vehicle speed of the tractor; through the cost function of the hydrogen energy composite power system, the dynamic predicted vehicle speed is subjected to rolling optimization processing under relevant output power to obtain the driving power requirement of the tractor; the driving power requirement power and the pre-acquired vehicle demand power are subjected to variational mode decomposition and reconstruction processing to obtain a multi-frequency power mode; based on the multi-frequency power mode, signal assignment processing is performed on the power battery, fuel cell and drive power generation system respectively; according to the real-time status of the battery SOC in the tractor, the hydrogen hybrid power is power redistributed to generate a real-time power distribution strategy.

[0009] The embodiments of the present application meet the requirements of green and pollution-free emission, long driving time, high system efficiency and good environmental adaptability by using fuel cells, power batteries and hydrogen internal combustion engine power generation drive systems. The designed energy management system is combined with the intelligent driving system and uses the environmental cognition model to assign time series speed predictions to achieve excellent adaptability to the environment and terrain. The collective power optimizer fully considers the multi-objective energy management of minimizing hydrogen consumption and long life of power batteries and fuel cells. At the same time, the power allocation strategy of multi-scale analysis fully combines the energy supply characteristics of fuel cells, power batteries and hydrogen internal combustion engines to enable them to operate in a high-efficiency range.

[0010] In a feasible implementation, multimodal data of the farmland environment is collected, and a data-driven farmland environment cognitive model is constructed based on key features of the farmland environment, specifically comprising: collecting whole-machine information of the tractor through a variety of sensors deployed in the tractor; wherein the whole-machine information includes at least: wheel-end torque, vehicle speed, traction and power demand; collecting farmland environment information; wherein the farmland environment information includes at least: soil type, slope and humidity; performing data fusion processing on the whole-machine information and the farmland environment information to obtain the multimodal data; and establishing an environment representation model based on the data mapping association in the multimodal data; extracting the key features of the farmland environment in the environment representation model, and performing data-driven matching on the key features of the farmland environment according to the reconstructed farmland ground to construct the farmland environment cognitive model; wherein the reconstructed farmland ground includes: flat road surface, gentle slope up, gentle slope down, steep slope up, steep slope down, large obstacles and small obstacles.

[0011] In a feasible implementation, the dynamic environmental features in the farmland environment cognitive model are extracted; based on the dynamic environmental features, the required speed of the tractor is calculated under the relevant environmental representation to obtain a required speed matrix under each environmental representation; through the TCN time series speed prediction model, and based on the required speed matrix, the historical speed information of the tractor is subjected to on-vehicle prediction processing within the prediction time interval to obtain an initial predicted speed value; according to Get the final predicted vehicle speed; where V final is the final predicted vehicle speed, V base is the historical vehicle speed, Δf i is the correction coefficient of environmental characteristics to vehicle speed, ω iThe environmental characterization model outputs a weight vector of each environmental feature in real time; N and i are both mathematical constants; a final predicted vehicle speed value predicted by the TCN neural network after environmental characterization correction and an actual vehicle speed value collected by a wheel speed sensor are obtained; a root mean square error evaluation is performed on the final predicted vehicle speed value and the actual vehicle speed value to obtain a vehicle speed prediction accuracy; and based on the vehicle speed prediction accuracy, the final predicted vehicle speed value is weighted and optimized to obtain the dynamic predicted vehicle speed.

[0012] In a feasible embodiment, the cost function of the hydrogen hybrid power system is used to perform rolling optimization processing on the dynamically predicted vehicle speed under the relevant output power to obtain the driving power requirement of the tractor, specifically including: determining a first power control amount corresponding to the dynamically predicted vehicle speed; performing decentralized calculations on the fuel cell hydrogen consumption, the SOC difference and the fuel cell change rate under different costs according to the fuel cell hydrogen consumption, the hydrogen internal combustion engine hydrogen consumption, the battery equivalent hydrogen consumption, the SOC reference value and the fuel cell change rate to generate the cost function; inputting the first power control amount into the cost function, and performing rolling optimization calculation on the first power control amount at the next moment based on the power control amount in the prediction time domain to predict and obtain the driving power requirement of the tractor.

[0013] In a possible embodiment, according to

[0014]

[0015] Get the cost function J; where C FC is the hydrogen consumption of the fuel cell, C ICE is the hydrogen consumption of the hydrogen internal combustion engine, C bat is the equivalent hydrogen consumption of the battery, SOC tar is the SOC reference value; ΔP FC is the fuel cell change rate; k and N are mathematical constants; α, β, and γ are weight factors under different costs; x k is the first power control amount; u k The power required for the driving.

[0016] In a feasible implementation, the driving demand power and the pre-acquired vehicle demand power are subjected to variational modal decomposition and reconstruction processing to obtain a multi-frequency power mode, specifically including: data fusion of the driving demand power and the vehicle demand power to obtain the load demand power; based on the power range limited by the high-efficiency interval of the energy components in the tractor, the load demand power is subjected to variational modal decomposition based on the frequency domain to obtain the multi-frequency power mode; wherein the multi-frequency power mode includes: low-frequency mode, medium-frequency mode and high-frequency mode.

[0017] In a feasible embodiment, based on the multi-frequency power mode, signal assignment processing is performed on the power battery, fuel cell and drive power generation system respectively, specifically including: transmitting and assigning the low-frequency power signal corresponding to the low-frequency mode to the fuel cell; wherein the steady-state power signal corresponds to the dynamic response of the fuel cell; transmitting and assigning the intermediate-frequency power signal corresponding to the intermediate-frequency mode to the control system of the drive power generation system; wherein the intermediate-frequency power signal corresponds to the basic load constant power output of the hydrogen internal combustion engine in the drive power generation system; transmitting and assigning the high-frequency power signal corresponding to the high-frequency mode to the control system of the power battery; wherein the high-frequency power signal corresponds to the transient compensation load of the power battery.

[0018] In a feasible embodiment, power is redistributed for the hydrogen hybrid power according to the real-time state of the battery SOC in the tractor, and a real-time power distribution strategy is generated, which specifically includes: performing real-time state monitoring on the battery SOC to obtain a battery state monitoring result; if the battery energy threshold of the battery state monitoring result is less than a first preset threshold, transferring part of the high-frequency components in the power battery to the fuel cell by adjusting the constraint conditions of variational mode decomposition (VMD) to obtain a first energy power distribution strategy; if the battery state monitoring result is a charging state, transferring the charging power load of the power battery to the real-time total power demand; based on variational mode decomposition (VMD), performing variational mode decomposition on the real-time total power demand, and distributing the decomposed real-time power load to the drive power generation system and the fuel cell.

[0019] In a feasible embodiment, the hydrogen hybrid power system mainly includes: a drive power generation system consisting of a hydrogen internal combustion engine and a generator, a fuel cell, a power battery, and a front and rear axle dual-acting clutch electric drive axle system; the drive power generation system is used to provide basic power for the tractor and charge the power battery; the fuel cell is used to provide the main energy source for the tractor; the power battery is used to store electrical energy and work in conjunction with the fuel cell at high power output to meet the instantaneous power requirements of the tractor; the front and rear axle dual-acting clutch electric drive axle system is used to distribute power output to the front and rear axles of the tractor according to different operating requirements.

[0020] On the other hand, an embodiment of the present application also provides a power system energy management device based on hydrogen hybrid power, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions that can be executed by the at least one processor, so that the at least one processor can execute a power system energy management method based on hydrogen hybrid power as described in any of the above embodiments.

[0021] This application provides a method and device for energy management of a power system based on hydrogen hybrid power. Compared with the prior art, the embodiments of this application have the following beneficial technical effects:

[0022] 1. Improve prediction accuracy: By collecting multimodal data on the farmland environment and building a cognitive model, the tractor's speed in different environments can be more accurately predicted, thereby improving the response speed and efficiency of the power system.

[0023] 2. Optimize vehicle speed control: Using the TCN (Temporal Convolutional Network) time series speed prediction model, accurate prediction of tractor speed can be achieved, which helps to optimize vehicle speed control and reduce energy consumption.

[0024] 3. Reduce costs: Through rolling optimization processing, the output power is adjusted according to the dynamic prediction of vehicle speed, which can make more efficient use of energy and reduce long-term operating costs.

[0025] 4. Power allocation optimization: Through variational mode decomposition, the power signal can be decomposed into different frequency components, thereby realizing a multi-scale power allocation strategy and improving the efficiency and performance of the power system.

[0026] 5. Improve system stability: Through precise power distribution strategy, the fluctuation of the power system under different working conditions can be reduced, and the stability and reliability of the system can be improved.

[0027] 6. Improve operating efficiency: Precise power system control can improve the operating efficiency of the tractor and reduce the extension of operating time caused by insufficient or excessive power consumption.

[0028] 7. Enhanced environmental adaptability: This method can adapt to different farmland environments and improve the tractor's operating performance in different environments by dynamically adjusting the power system.

[0029] 8. Reduce energy consumption: By optimizing power distribution and vehicle speed control, energy consumption can be reduced, meeting the requirements of energy conservation and emission reduction. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments described in the present application. For those skilled in the art, other drawings can be obtained based on these drawings without creative work. In the drawings:

[0031] Figure 1 A flow chart of a method for energy management of a power system based on hydrogen hybrid power provided in an embodiment of the present application;

[0032] Figure 2 A schematic diagram of a hydrogen hybrid tractor configuration provided in an embodiment of the present application; wherein: 1 - hydrogen internal combustion engine, 2 - wet clutch, 3 - wet clutch, 4 - drive motor, 5 - power battery, 6 - drive motor, 7 - double-acting clutch, 8 - PTO shaft, 9 - PTO shaft drive motor, 10 - fuel cell, 11 - generator;

[0033] Figure 3 A flow chart of speed planning control based on farmland environment cognition provided in an embodiment of the present application;

[0034] Figure 4 A schematic structural diagram of an energy management device for a power system based on hydrogen hybrid power provided in an embodiment of the present application. DETAILED DESCRIPTION

[0035] In order to enable those skilled in the art to better understand the technical solutions in this application, the following will clearly and completely describe the technical solutions in the embodiments of this application in conjunction with the drawings in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments of this specification, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of this application.

[0036] It should be noted that Figure 2A schematic diagram of a hydrogen hybrid tractor configuration provided in an embodiment of the present application is shown in FIG. Figure 2 As shown, the hydrogen hybrid tractor configuration consists of a drive and power generation system consisting of a hydrogen internal combustion engine 1 and a generator 11, a fuel cell 10, a power battery 5, and an electric drive axle with a dual-acting clutch 7 on the front and rear axles. The drive and power generation system composed of the hydrogen internal combustion engine 1 and generator 11 provides basic power for the tractor and recharges the power battery when needed. The use of the hydrogen internal combustion engine 1 not only reduces the pollution emissions of traditional diesel engines but also improves energy efficiency. The fuel cell 10 serves as the primary energy source, providing continuous power support for the tractor. The power battery 5 stores electrical energy and works in conjunction with the fuel cell when high power output is required to meet the tractor's instantaneous power requirements. When the tractor is performing rotary tillage operations, the PTO shaft drive motor 9 cooperates with the dual-acting clutch 7 on the rear axle to determine whether coupled power output is required based on the power requirements of the rotary tillage operation. This design ensures that the tractor can provide appropriate power output under different operating conditions, improving operating efficiency. The dual-acting clutch 7 on the front and rear axles and the electric drive axle design allows the tractor to flexibly allocate power output between the front and rear axles according to different operating requirements. When four-wheel drive is required, it can provide stronger traction; when energy conservation is needed, it can drive only a single axle, improving energy efficiency. The designed external power socket enables the tractor to function as a mobile power station. This not only facilitates charging other equipment in the field but also increases the tractor's versatility, allowing it to be used even during off-season. Through intelligent energy management strategies, power distribution between the hydrogen internal combustion engine 1, generator 11, fuel cell 10, and power battery 5 is optimized based on the tractor's real-time needs and the state of the energy source. This improves overall system efficiency, reduces energy waste, and extends battery life. This configuration adapts well to the tractor's needs in various agricultural operations, such as tilling, harvesting, and transport. Furthermore, the system's modular design facilitates future upgrades and expansion. It includes a hydrogen hybrid power system with a PTO shaft power unit integrated into the power system to meet both driving and tilling power requirements. Sensors such as lidar, GNSS, an inertial measurement unit, and a multispectral camera are deployed, and the fuel cell can be a hydrogen fuel cell.

[0037] The existing general tractor power working modes include: pure electric traction: power battery + fuel cell + generator, which has a higher SOC requirement for the power battery; hydrogen internal combustion engine hybrid: hydrogen internal combustion engine drive + power generation / power battery + fuel cell combination, which has a lower SOC requirement for the power battery; mobile power station: power battery + fuel cell, which requires external plug-in; PTO output: dual motor drive for rotary tillage.

[0038] The embodiment of the present application provides a method for energy management of a power system based on hydrogen hybrid power, such as Figure 1 As shown, the power system energy management method based on hydrogen hybrid power specifically includes steps S101-S106:

[0039] S101. Collect multimodal data of farmland environment and build a data-driven farmland environment cognitive model based on key features of the farmland environment.

[0040] Specifically, various sensors deployed in the tractor are first used to collect the tractor's overall information, which includes at least wheel-end torque, vehicle speed, traction, and power requirements.

[0041] Furthermore, farmland environmental information is collected; wherein the farmland environmental information at least includes: soil type, slope and humidity.

[0042] Furthermore, the whole machine information and farmland environment information are fused to obtain multimodal data; and based on the data mapping association in the multimodal data, an environmental representation model is established.

[0043] In one embodiment, to ensure high efficiency and system durability under typical conditions such as plowing, rotary tillage, and transport, an energy management strategy is designed that integrates speed planning control and multi-objective power reconstruction with local environmental perception data. By deploying torque sensors, wheel speed sensors, and other operating status sensors such as GPS and an inertial measurement unit on the tractor, key operating parameters of the power system (such as wheel-end torque, vehicle speed, traction, and power demand) are collected during operation. This data is then combined with farmland environmental information (such as soil type, slope, and moisture) for data fusion and processing to establish an environmental representation model.

[0044] Furthermore, the key features of the farmland environment are extracted from the environmental representation model and matched with the reconstructed farmland surface through data-driven matching to construct a farmland environmental cognitive model. The reconstructed farmland surface includes: flat road surface, gentle uphill slope, gentle downhill slope, steep uphill slope, steep downhill slope, large obstacles, and small obstacles.

[0045] As a feasible implementation method, Figure 3 The present application provides a flow chart of speed planning control based on farmland environment cognition, such as Figure 3 As shown, deep learning algorithms and data fusion techniques are used to process and analyze real-time multimodal farmland environmental data, including vehicle location, speed, crop distribution, and road conditions. Key features of the farmland environment are extracted and a data-driven farmland environmental cognitive model is constructed. Furthermore, to reduce the computing power required by the controller, the farmland surface is reconstructed into seven states: flat road, gentle upslope, gentle downslope, steep upslope, steep downslope, large obstacles, and small obstacles, thereby reducing algorithm complexity.

[0046] S102. According to the dynamic environmental characteristics in the farmland environment cognitive model and based on the TCN time series speed prediction model, the tractor's speed is predicted under the relevant environmental characteristics to obtain the dynamic predicted speed of the tractor.

[0047] Specifically, it is necessary to first extract the dynamic environmental characteristics in the farmland environmental cognitive model.

[0048] Furthermore, according to the dynamic environmental characteristics, the required speed of the tractor under the relevant environmental representation is calculated based on the vehicle speed information, and the required speed matrix under each environmental representation is obtained.

[0049] Furthermore, the TCN time series speed prediction model is used to perform on-vehicle prediction processing on the tractor's historical speed information within the prediction time interval based on the required speed matrix to obtain the initial predicted speed value.

[0050] Further, according to Get the final predicted vehicle speed; where V final is the final predicted vehicle speed, V base is the historical vehicle speed, Δf i is the correction coefficient of environmental characteristics to vehicle speed, ω i The environmental characterization model outputs the weight vector of each environmental feature in real time; N and i are both mathematical constants.

[0051] Furthermore, the final predicted vehicle speed value predicted by the TCN neural network after environmental characterization correction and the actual vehicle speed value collected by the wheel speed sensor are obtained.

[0052] Furthermore, the final predicted vehicle speed value and the actual vehicle speed value are subjected to a root mean square error evaluation process to obtain the vehicle speed prediction accuracy.

[0053] Furthermore, according to the vehicle speed prediction accuracy, the final predicted vehicle speed value is weighted and optimized to obtain a dynamic predicted vehicle speed.

[0054] In one embodiment, Figure 3 As shown in the figure, the established TCN time series speed prediction model is first used to adjust the vehicle speed according to the dynamic environmental characteristics in the farmland environment cognitive model to cope with the upcoming environmental mutation. In the future control time domain T, the historical vehicle speed information at the past moment is used as input, and every time t k To predict a point, the TCN neural network power prediction process is as follows: select the current time t and the historical vehicle speed information of the past k time as input: V his =V t-k , V t-k+1 ,…,V t , then output the speed of the vehicle at k time in the future: V fut=f V (V his )=V t+1 , V t+2 ,…,V t+k Among them, V his Indicates historical vehicle speed information, V fut Represents the predicted vehicle speed information output by the TCN neural network. The TCN model learns the mapping relationship between input data and output data f V This mapping relationship is automatically learned through the model training process, usually using historical data for supervised learning. k is the prediction time. The output weight ω of the environment representation is given i The final predicted vehicle speed within the prediction time domain T is obtained.

[0055] In one embodiment, the root mean square error (RMSE), which is the square root of the ratio of the square of the deviation between the vehicle speed value predicted by the TCN neural network and the actual vehicle speed value collected by the wheel speed sensor and the number of observations, is also used to evaluate the vehicle speed prediction accuracy of the TCN neural network, namely: That is, the smaller the RMSE, the higher the prediction accuracy. If the accuracy is too low, the weights are readjusted to optimize the vehicle speed, and finally the dynamic predicted vehicle speed is optimized.

[0056] S103 , performing rolling optimization processing on the dynamically predicted vehicle speed under the relevant output power using the cost function of the hydrogen energy hybrid power system to obtain the required driving power of the tractor.

[0057] Specifically, the driving demand power and the vehicle demand power are first fused to obtain the load demand power, and then the first power control amount corresponding to the dynamically predicted vehicle speed is determined, that is, the first control amount in the predicted time domain of the dynamically predicted vehicle speed is used as the input at this moment.

[0058] Furthermore, based on the hydrogen consumption of the fuel cell, the hydrogen consumption of the hydrogen internal combustion engine, the equivalent hydrogen consumption of the battery, the SOC reference value and the fuel cell change rate, the fuel cell hydrogen consumption, the SOC difference and the fuel cell change rate are all decentralized calculated under different costs to generate a cost function for the hydrogen hybrid power system.

[0059] Furthermore, the first power control amount is input into the cost function, and based on the power control amount in the prediction time domain, a rolling optimization calculation is performed on the first power control amount at the next moment to predict and obtain the required driving power of the tractor.

[0060] As a feasible implementation method, in order to improve the hydrogen fuel economy of the thermoelectric hybrid power system as much as possible and to extend the service life of the lithium battery pack and fuel cell system as much as possible, a power optimizer is constructed to optimize the output power based on the cost function of minimum hydrogen consumption and long life of the power battery and fuel cell. The cost function is composed of the following formula: the first power control quantity obtained in the predicted time domain is used as the input at this moment, and the control quantity at the next moment is optimized online using the same method. The optimization is carried out in a rolling manner until it is terminated to obtain the optimal output power (driving demand power). That is, it is necessary to

[0061] Get the cost function J. Among them, C FC is the hydrogen consumption of the fuel cell, C ICE is the hydrogen consumption of hydrogen internal combustion engine, C bat is the equivalent hydrogen consumption of the battery, SOC tar It is the SOC reference value, and the purpose of setting this item is to prevent overcharge and overdischarge, thereby protecting the lithium battery; ΔP FC is the fuel cell change rate, and the fuel cell change rate ΔP is set FC The absolute value of is to prevent the fuel cell power from increasing or decreasing suddenly and prolong its service life; k and N are mathematical constants; α, β, and γ are weight factors under different costs; x k is the first power control amount; u k Power is required for driving.

[0062] S104 , performing variational mode decomposition and reconstruction processing on the driving demand power and the pre-acquired vehicle demand power to obtain a multi-frequency power mode.

[0063] Specifically, the driving power requirement and the vehicle power requirement are first fused to obtain the load power requirement. Then, based on the high-efficiency range of the tractor's energy components, the load power requirement is subjected to a frequency-domain variational modal decomposition to obtain multi-frequency power modes. These multi-frequency power modes include low-frequency modes, mid-frequency modes, and high-frequency modes.

[0064] In one embodiment, the mathematical model of variational mode decomposition is as follows where u k (t) is the kth mode, corresponding to the power sub-requirements of different frequency characteristics, and P(t) is the total required power, and then the low-frequency power signal u1(t) is assigned to the fuel cell, the medium-frequency power signal u2(t) is assigned to the drive power generation system composed of the hydrogen internal combustion engine and the generator, and the high-frequency power u3(t) is assigned to the power battery.

[0065] S105 , based on the multi-frequency power mode, respectively perform signal assignment processing on the power battery, fuel cell, and drive power generation system.

[0066] Specifically, the steady-state power signal corresponding to the low-frequency mode is transmitted and assigned to the fuel cell, wherein the low-frequency power signal corresponds to the base load constant power output of the fuel cell.

[0067] Furthermore, the intermediate frequency power signal corresponding to the intermediate frequency mode is transmitted and assigned to the control system of the driving power generation system. The intermediate frequency power signal corresponds to the dynamic response of the fuel cell.

[0068] Furthermore, a high-frequency power signal corresponding to the high-frequency mode is transmitted and assigned to the control system of the power battery, wherein the high-frequency power signal corresponds to the transient compensation load of the power battery.

[0069] In one implementation, the power range is limited by the high-efficiency interval of the energy component, and the variational mode decomposition (VMD) method is used to decompose the required power in the frequency domain, specifically including: VMD output: 3 modes, corresponding to low, medium and high frequency components, low-frequency mode (P1), that is, when it is in a low-frequency power signal: <0.01Hz, corresponding to the fuel cell power output. Medium-frequency mode (P2), that is, when it is in a medium-frequency power signal: 0.01~0.1Hz, corresponding to the power output of the hydrogen internal combustion engine that drives the power generation system. High-frequency mode (P3), that is, when it is in a high-frequency power signal: >0.1Hz, corresponding to the transient compensation load of the power battery.

[0070] S106: Redistribute power of the hydrogen hybrid power according to the real-time state of the battery SOC in the tractor, and generate a real-time power distribution strategy.

[0071] Specifically, the battery SOC is monitored in real time to obtain a battery status monitoring result.

[0072] Furthermore, if the battery energy threshold of the battery status monitoring result is less than the first preset threshold, part of the high-frequency components in the power battery are transferred to the fuel cell by adjusting the constraint conditions of the variational mode decomposition (VMD), thereby obtaining a first energy and power allocation strategy.

[0073] Furthermore, if the battery status monitoring result indicates that charging is enabled, the charging power load of the power battery is transferred to the real-time total power demand. Based on variational mode decomposition (VMD), the real-time total power demand is decomposed into variational modes and the decomposed real-time power load is distributed to the drive power generation system and fuel cell.

[0074] As a feasible implementation method, the hydrogen hybrid power system in the hydrogen hybrid tractor configuration of the present application mainly includes: a drive power generation system composed of a hydrogen internal combustion engine and a generator, a fuel cell, a power battery, and a front and rear axle dual-acting clutch electric drive axle system. The drive power generation system is used to provide basic power for the tractor and charge the power battery. The fuel cell is used to provide the main energy source for the tractor. The power battery is used to store electrical energy and work in conjunction with the fuel cell at high power output to achieve the instantaneous power demand of the tractor. The front and rear axle dual-acting clutch electric drive axle system is used to distribute the power output of the tractor to the front and rear axles according to different operating requirements.

[0075] At the same time, if Figure 2 As shown, based on the design of the hydrogen hybrid intelligent tractor, during rotary tillage operations, the PTO shaft drive motor and the rear axle's double-acting clutch determine whether to couple power output based on power requirements, ensuring appropriate power output under different operating conditions. The rear axle's double-acting clutch electric drive axle flexibly adjusts power output according to operational needs, enabling four-wheel drive or single-axle drive, improving energy efficiency. An external power outlet: The designed external power outlet enables the tractor to function as a mobile power station, conveniently charging other equipment in the field and increasing the tractor's versatility.

[0076] In addition, the embodiment of the present application also provides a power system energy management device based on hydrogen hybrid power, such as Figure 4 As shown, the power system energy management device 400 based on hydrogen hybrid power specifically includes:

[0077] At least one processor 401; and a memory 402 in communication with the at least one processor 401; wherein the memory 402 stores instructions executable by the at least one processor 401, so as to enable the at least one processor 401 to perform:

[0078] Collect multimodal data of farmland environment and build a data-driven farmland environment cognitive model based on key features of the farmland environment;

[0079] According to the dynamic environmental characteristics in the farmland environment cognitive model and based on the TCN time series speed prediction model, the tractor's speed is predicted under the relevant environmental characteristics to obtain the dynamic predicted speed of the tractor;

[0080] Through the cost function of the hydrogen hybrid power system, the dynamic predicted vehicle speed is subjected to rolling optimization processing under the relevant output power to obtain the required driving power of the tractor.

[0081] Decomposition and reconstruction of the driving demand power and the pre-acquired vehicle demand power are performed on the variational mode to obtain a multi-frequency power mode;

[0082] Based on multi-frequency power mode, signal assignment processing is performed on the power battery, fuel cell and drive power generation system respectively;

[0083] According to the real-time state of battery SOC in the tractor, power is redistributed for the hydrogen hybrid and a real-time power allocation strategy is generated.

[0084] The embodiments of the present application meet the requirements of green and pollution-free emission, long driving time, high system efficiency and good environmental adaptability by using fuel cells, power batteries and hydrogen internal combustion engine power generation drive systems. The designed energy management system is combined with the intelligent driving system and uses the environmental cognition model to assign time series speed predictions to achieve excellent adaptability to the environment and terrain. The collective power optimizer fully considers the multi-objective energy management of minimizing hydrogen consumption and long life of power batteries and fuel cells. At the same time, the power allocation strategy of multi-scale analysis fully combines the energy supply characteristics of fuel cells, power batteries and hydrogen internal combustion engines to enable them to operate in a high-efficiency range.

[0085] The various embodiments in this application are described in a progressive manner. Similar parts between the various embodiments can be referred to in conjunction with each other. Each embodiment focuses on the differences between the other embodiments. In particular, the device embodiments are generally similar to the method embodiments, so the description is relatively simple. For relevant parts, refer to the description of the method embodiments.

[0086] The foregoing description describes specific embodiments of the present application. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims can be performed in an order different from that described in the embodiments and still achieve the desired results. Furthermore, the processes depicted in the accompanying drawings do not necessarily require the specific order shown or the sequential order to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0087] The foregoing is merely an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the embodiments of the present application may have various modifications and variations. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the embodiments of the present application should be included within the scope of the specification of the present application.

Claims

1. A method for energy management of a power system based on hydrogen hybrid power, characterized in that: The method comprises: Collect multimodal data of farmland environment and build a data-driven farmland environment cognitive model based on key features of the farmland environment; According to the dynamic environmental characteristics in the farmland environment cognitive model and based on the TCN time series speed prediction model, the tractor's speed is predicted under the relevant environmental characteristics to obtain the dynamic predicted speed of the tractor; The dynamic predicted vehicle speed is subjected to rolling optimization processing under the relevant output power through the cost function of the hydrogen energy hybrid power system to obtain the required driving power of the tractor; Decomposing and reconstructing the driving demand power and the pre-acquired vehicle demand power in a variational mode to obtain a multi-frequency power mode; Based on the multi-frequency power mode, signal assignment processing is performed on the power battery, fuel cell and drive power generation system respectively; According to the real-time state of the battery SOC in the tractor, power is redistributed for the hydrogen hybrid power to generate a real-time power distribution strategy.

2. The method for energy management of a power system based on hydrogen hybrid power according to claim 1, characterized in that: Collect multimodal data of farmland environment and build a data-driven farmland environment cognitive model based on key features of farmland environment, including: Collecting the tractor's overall information through various sensors deployed in the tractor; wherein the overall information includes at least: wheel end torque, vehicle speed, traction, and power demand; Collecting farmland environmental information; wherein the farmland environmental information includes at least: soil type, slope and humidity; Performing data fusion processing on the whole machine information and the farmland environment information to obtain the multimodal data; and establishing an environmental representation model based on data mapping association in the multimodal data; The key features of the farmland environment in the environmental characterization model are extracted, and according to the reconstructed farmland ground, the key features of the farmland environment are data-driven matched to construct the farmland environment cognitive model; wherein the reconstructed farmland ground includes: flat road surface, gentle uphill slope, gentle downhill slope, steep uphill slope, steep downhill slope, large obstacles and small obstacles.

3. The method for energy management of a power system based on hydrogen hybrid power according to claim 1, characterized in that: According to the dynamic environmental characteristics in the farmland environment cognitive model and based on the TCN time series speed prediction model, the tractor's speed is predicted under the relevant environmental characteristics to obtain the dynamic predicted speed of the tractor, specifically including: extracting dynamic environmental features in the farmland environmental cognitive model; According to the dynamic environmental characteristics, the required vehicle speed of the tractor is calculated under the relevant environmental representations to obtain a required vehicle speed matrix under each environmental representation; Using the TCN time series speed prediction model and based on the required speed matrix, the historical speed information of the tractor is subjected to on-vehicle prediction processing within a prediction time interval to obtain an initial predicted speed value; according to Get the final predicted vehicle speed; where V final is the final predicted vehicle speed, V base is the historical vehicle speed, Δf i is the correction coefficient of environmental characteristics to vehicle speed, ω i The environmental representation model outputs the weight vector of each environmental feature in real time; N and i are both mathematical constants; Obtain the final predicted vehicle speed value predicted by the TCN neural network after environmental characterization correction and the actual vehicle speed value collected by the wheel speed sensor; Performing a root mean square error evaluation process on the final predicted vehicle speed value and the actual vehicle speed value to obtain a vehicle speed prediction accuracy; According to the vehicle speed prediction accuracy, the final predicted vehicle speed value is weighted and optimized to obtain the dynamic predicted vehicle speed.

4. The method for energy management of a power system based on hydrogen hybrid power according to claim 1, characterized in that: The dynamic predicted vehicle speed is subjected to rolling optimization processing under the output power by using the cost function of the hydrogen hybrid power system to obtain the required driving power of the tractor, specifically including: determining a first power control amount corresponding to the dynamically predicted vehicle speed; According to the hydrogen consumption of the fuel cell, the hydrogen consumption of the hydrogen internal combustion engine, the equivalent hydrogen consumption of the battery, the SOC reference value and the fuel cell change rate, the fuel cell hydrogen consumption, the SOC difference and the fuel cell change rate are all subjected to weighted calculation under different costs to generate the cost function; The first power control amount is input into the cost function, and based on the power control amount in the prediction time domain, a rolling optimization calculation is performed on the first power control amount at the next moment to predict and obtain the required driving power of the tractor.

5. The method for energy management of a power system based on hydrogen hybrid power according to claim 4, characterized in that: according to Get the cost function J; where C FC is the hydrogen consumption of the fuel cell, C ICE is the hydrogen consumption of the hydrogen internal combustion engine, C bat is the equivalent hydrogen consumption of the battery, SOC tar is the SOC reference value; ΔP FC is the fuel cell change rate; k and N are mathematical constants; α, β, and γ are weight factors under different costs; x k is the first power control amount; u k The power required for the driving.

6. The method for energy management of a power system based on hydrogen hybrid power according to claim 1, characterized in that: The driving demand power and the pre-acquired vehicle demand power are subjected to variational mode decomposition and reconstruction processing to obtain a multi-frequency power mode, specifically including: Fusing the driving power requirement with the vehicle power requirement to obtain the load power requirement; Based on the power range limited by the high-efficiency interval of the energy components in the tractor, the load demand power is subjected to variational modal decomposition based on the frequency domain to obtain the multi-frequency power mode; wherein the multi-frequency power mode includes: low-frequency mode, medium-frequency mode and high-frequency mode.

7. The method for energy management of a power system based on hydrogen hybrid power according to claim 6, characterized in that: Based on the multi-frequency power mode, signal assignment processing is performed on the power battery, fuel cell, and drive power generation system respectively, specifically including: Transmitting and assigning a low-frequency power signal corresponding to the low-frequency mode to the fuel cell; wherein the low-frequency power signal corresponds to the base load constant power output of the fuel cell; Transmitting and assigning an intermediate frequency power signal corresponding to the intermediate frequency mode to a control system driving a power generation system; wherein the intermediate frequency power signal corresponds to a dynamic response of a hydrogen internal combustion engine; The high-frequency power signal corresponding to the high-frequency mode is transmitted and assigned to the control system of the power battery; wherein the high-frequency power signal corresponds to the transient compensation load of the power battery.

8. The method for energy management of a power system based on hydrogen hybrid power according to claim 1, characterized in that: According to the real-time state of the battery SOC in the tractor, power is redistributed for the hydrogen hybrid power, and a real-time power distribution strategy is generated, specifically including: Performing real-time status monitoring on the battery SOC to obtain a battery status monitoring result; If the battery energy threshold of the battery state monitoring result is less than a first preset threshold, a first energy and power allocation strategy is obtained by transferring part of the high-frequency components in the power battery to the fuel cell by adjusting the constraint conditions of the variational mode decomposition (VMD); If the battery status monitoring result is that the charging state is turned on, the charging power load of the power battery is transferred to the real-time total power demand; based on variational mode decomposition VMD, the real-time total power demand is decomposed into variational modes, and the decomposed real-time power load is distributed to the drive power generation system and the fuel cell.

9. The method for energy management of a power system based on hydrogen hybrid power according to claim 1, characterized in that: The hydrogen hybrid system mainly includes: a drive and power generation system consisting of a hydrogen internal combustion engine and a generator, a fuel cell, a power battery, and a front and rear axle double-acting clutch electric drive axle system; The driving power generation system is used to provide basic power for the tractor and charge the power battery; The fuel cell is used to provide a main energy source for the tractor; The power battery is used to store electrical energy and work in conjunction with the fuel cell during high power output to meet the instantaneous power demand of the tractor; The front and rear axle dual-acting clutch electric drive axle system is used to distribute power output to the front and rear axles of the tractor according to different operating requirements.

10. A power system energy management device based on hydrogen hybrid power, characterized in that: The device comprises: at least one processor; and, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, so that the at least one processor can execute the energy management method of a power system based on hydrogen hybrid power according to any one of claims 1 to 9.

Images