Hydrogen energy heavy truck multi-mode power system parameter matching method

Abstract

The application belongs to the technical field of hydrogen energy heavy truck power system, and specifically discloses a hydrogen energy heavy truck multi-mode power system parameter matching method, which comprises the following steps: constructing a power system component model, obtaining a vehicle power demand time sequence of a typical working condition, taking the vehicle power demand time sequence as a constraint input for capacity matching and energy flow reconstruction, and using a mixed integer two-stage optimization algorithm to perform collaborative optimization on the capacity parameters and topology switching strategy of the multi-mode power system. The hydrogen energy heavy truck multi-mode power system parameter matching method can significantly suppress energy consumption fluctuation under frequent start-stop and complex road conditions, delay the aging rate of core components, and comprehensively improve the economy, reliability and adaptability of the hydrogen energy heavy truck multi-mode power system.

Description

Technical Field

[0001] This invention belongs to the technical field of hydrogen-powered heavy-duty truck power systems, specifically relating to a method for matching parameters of a multi-mode power system for hydrogen-powered heavy-duty trucks. Background Technology

[0002] With the proposal and promotion of the "dual carbon" goal, hydrogen energy, due to its clean, efficient, and renewable characteristics, has become an important development direction for heavy-duty trucks to replace traditional diesel power. Currently, hydrogen fuel cell heavy-duty trucks have been demonstrated in urban logistics and short-distance transportation, but under typical operating conditions such as long-distance and high-load conditions, the driving range and power response relying solely on fuel cells are still insufficient to meet the demands. To address this, the industry has begun to explore multi-mode power system architectures such as "fuel cell + energy storage battery + auxiliary engine / supercapacitor," achieving an organic combination of high power output and long driving range by switching or paralleling various subsystems under different operating conditions.

[0003] However, existing multi-mode power systems still have many shortcomings in practical applications: First, there is a lack of systematic methods for matching the capacity of fuel cell stacks and energy storage batteries, which often leads to decreased efficiency and increased cost and weight under low load conditions; Second, mode switching strategies are mostly based on experience to set thresholds, without fully considering the dynamic efficiency curves of each subsystem, which can easily cause energy consumption fluctuations under frequent start-stop or complex road conditions. Summary of the Invention

[0004] The purpose of this invention is to solve the problem of matching energy efficiency and power performance of multi-mode power systems for hydrogen-powered heavy trucks under complex operating conditions, and to propose a parameter matching method for multi-mode power systems for hydrogen-powered heavy trucks.

[0005] The technical solution of this invention is as follows: This invention provides a method for parameter matching of a multi-mode power system for hydrogen-powered heavy-duty trucks, comprising: Step 1: Construct power system component models for the multi-mode power system of hydrogen-powered heavy trucks and set up online model self-calibration strategies; Step 2: Collect the real-time power demand of the vehicle, determine the typical operating conditions of the vehicle based on the real-time power demand, and then splice the power demand subsequences under each typical operating condition to obtain the vehicle power demand time series. The vehicle power demand time series serves as the constraint input for capacity matching and energy flow reconstruction. Step 3: Use a mixed integer two-stage optimization algorithm to perform collaborative optimization of the capacity parameters and topology switching strategy of the multi-mode power system. The first stage of the collaborative optimization determines the capacity parameters with the vehicle power demand time series as the constraint boundary. The second stage uses the switching state of the reconfigurable DC bus converter of the multi-mode power system as the integer decision variable to generate a topology switching calendar that matches the typical operating conditions corresponding to the vehicle power demand time series. Step 4: Input the capacity parameters and the topology switching calendar into the vehicle simulation test platform for performance verification. If the simulation results do not meet the preset target, adjust the optimization weight coefficients of the mixed integer two-stage optimization algorithm and re-execute step 3 until the simulation results meet the preset target.

[0006] Preferably, the power system component model includes a fuel cell stack, an energy storage battery pack, a supercapacitor, and a reconfigurable DC bus converter.

[0007] Preferably, the online model self-calibration strategy includes automatically updating the efficiency curves of the fuel cell stack, the energy storage battery pack, and the supercapacitor every 10,000 km of cumulative vehicle mileage to compensate for model errors caused by component performance degradation.

[0008] Preferably, step 2 includes: Step 2.1: Based on historical vehicle operation data, cluster analysis is performed to identify several typical operating conditions; Step 2.2: Collect the real-time power demand of the vehicle, and determine the typical operating conditions corresponding to the vehicle based on the real-time power demand; Step 2.3: Generate power demand subsequences for each typical operating condition by weighted fusion of data collected by the on-board data recorder and bench simulation data; Step 2.4: Concatenate the power demand subsequences according to the time sequence logic of the actual vehicle driving to obtain the vehicle power demand time sequence.

[0009] Preferably, the capacity parameters include the rated power of the fuel cell stack, the capacity of the energy storage battery pack, and the capacity of the supercapacitor.

[0010] Preferably, the mixed-integer two-stage optimization algorithm uses a multi-objective function J as the optimization objective, and the expression of the multi-objective function J is: ; in, Indicates time Vehicle power requirements For system efficiency, For time step, Indicates the mass of the fuel cell stack. Indicates the mass of the energy storage battery pack. Indicates the mass of a supercapacitor. This indicates the power density of the fuel cell stack. This indicates the power density of the energy storage battery pack. This indicates the power density of a supercapacitor. Indicates the capacity of the fuel cell stack. Indicates the capacity of the energy storage battery pack. This indicates the capacitance of the supercapacitor. This indicates the energy density of the fuel cell stack. This indicates the energy density of the energy storage battery pack. This indicates the energy density of a supercapacitor. The first optimized weight coefficient, For the second optimization weighting coefficient, For the third optimization weight coefficient, Indicates time, This represents the total number of time steps.

[0011] Preferably, during the optimization process of the multi-objective function J, upper limits of the energy storage battery health status and upper limits of the fuel cell voltage decay are set. When either upper limit constraint is triggered, the second optimization weight coefficient is increased.

[0012] Preferably, the switching state of the reconfigurable DC bus converter of the multi-mode power system is used as an integer decision variable to generate a topology switching calendar that matches the typical operating conditions corresponding to the vehicle power demand time series, including: Based on the determined capacity parameters, the energy flow topology state of the multi-mode power system is selected by taking the switching state of the reconfigurable DC bus converter of the multi-mode power system as an integer decision variable. The energy flow topology state is jointly modeled with the continuous power allocation variable, and the operating parameters under each energy flow topology state are calculated. The operating parameters include the power threshold and the energy storage battery state of charge window. The continuous time periods that meet the operating parameters are determined segment by segment along the time axis of the vehicle power demand time series. The topology switching calendar is generated based on the energy flow topology state and start and end times associated with each continuous time period. The switching cost corresponding to the topology switching calendar is calculated based on the topology switching penalty function, and the topology switching calendar is output with the goal of minimizing the switching cost.

[0013] Preferably, the expression for the topology switching penalty function is: ; in, This represents the number of topology switches within the topology switch calendar period. This represents the energy loss during the i-th switching process. Let be the power interruption time caused by the i-th switching. This represents the average drive power within the 30 seconds prior to the switch.

[0014] Preferably, the preset targets include hydrogen consumption per 100 kilometers and acceleration time.

[0015] The beneficial effects of this invention are: The parameter matching method for multi-mode power systems of hydrogen-powered heavy-duty trucks of this invention utilizes an integrated "capacity-topology" matching mechanism to collaboratively determine the rated power of the fuel cell stack, the capacity of the power battery pack, and the capacity of the supercapacitor within the same mixed-integer two-stage optimization framework. Combined with topology adaptive reconfiguration of the reconfigurable DC bus converter, this method can significantly reduce energy loss and redundancy under low-load conditions while meeting peak power demands. Compared to existing solutions that rely on experience-based selection, this fundamentally solves the problems of efficiency degradation and excessive cost caused by the lack of a systematic approach to matching the capacity of fuel cell stacks and energy storage devices. Furthermore, it can significantly suppress energy consumption fluctuations under frequent start-stop cycles and complex road conditions, and slow down the aging rate of core components, comprehensively improving the economy, reliability, and adaptability of multi-mode power systems for hydrogen-powered heavy-duty trucks. Attached Figure Description

[0016] Figure 1 The flowchart shown is a method for parameter matching of a multi-mode power system for hydrogen-powered heavy trucks; Figure 2 The image shown is a schematic diagram of a power system component model; Figure 3 The diagram shows a flowchart of the process for obtaining vehicle power demand time series data. Figure 4 The diagram shows a schematic of the system's collaborative optimization process. Detailed Implementation

[0017] Exemplary embodiments of the present invention will now be described in detail with reference to the accompanying drawings. It should be understood that the embodiments shown and described in the drawings are merely exemplary and are intended to illustrate the principles and spirit of the invention, and are not intended to limit the scope of the invention.

[0018] This invention provides a method for matching parameters of a multi-mode power system for hydrogen-powered heavy-duty trucks. Please refer to [link / reference]. Figure 1 , Figure 1 The diagram shows a flowchart of a method for matching parameters of a multi-mode power system for hydrogen-powered heavy trucks.

[0019] like Figure 1 As shown, the parameter matching method for the multi-mode power system of hydrogen-powered heavy-duty trucks in this embodiment includes the following steps: Step 1: Construct power system component models for the multi-mode power system of hydrogen-powered heavy trucks and set up online model self-calibration strategies.

[0020] Please see Figure 2 , Figure 2 The image shown is a schematic diagram of a power system component model, as follows: Figure 2As shown, in this embodiment, the power system component model includes a fuel cell stack, an energy storage battery pack, a supercapacitor, and a reconfigurable DC bus converter. Each power system component model is used to record the efficiency curves, mass, volume, and cost parameters of each component.

[0021] Understandably, the established complete powertrain component model includes the performance parameters of each component in the vehicle's powertrain. Specifically, the fuel cell stack provides the core power source, determining the vehicle's main power output. The energy storage battery pack stores electrical energy, providing additional power to the vehicle, especially during acceleration or when power demands are high. Supercapacitors are used to rapidly release or absorb energy, optimizing the response to short-term power fluctuations and improving the system's dynamic performance.

[0022] In this embodiment, the online model self-calibration strategy includes automatically updating the efficiency curves of the fuel cell stack, energy storage battery pack, and supercapacitor every 10,000 km of cumulative vehicle mileage to compensate for model errors caused by component performance degradation.

[0023] For example, an 8×4 hydrogen fuel cell tractor with a total mass of 49t is used as the subject. The initial power system component model is as follows: the fuel cell stack has a rated net power of 300kW, corresponding to a mass of approximately 450kg, a volume of approximately 310L, and a unit price of approximately 580,000 yuan; the energy storage battery pack has a nominal energy of 120kWh, a mass of approximately 780kg, a volume of approximately 390L, and a unit price of approximately 420,000 yuan; the supercapacitor has an equivalent energy of 5MJ, a mass of approximately 165kg, a volume of approximately 95L, and a unit price of approximately 90,000 yuan; the reconfigurable DC bus converter adopts two 250kW bidirectional DC / DC modules, with a total mass of approximately 95kg, a volume of approximately 45L, and a unit price of approximately 110,000 yuan.

[0024] The specific calibration process of the initial efficiency curve is as follows: A stepped power bench test was conducted at 25℃, and the net efficiency at three load points of 100kW, 200kW and 300kW was recorded respectively: the corresponding values ​​for the fuel cell stack were 0.605, 0.558 and 0.514; the discharge channel efficiency of the energy storage battery was 0.941, 0.931 and 0.917; and the discharge channel efficiency of the supercapacitor was 0.967, 0.960 and 0.953.

[0025] Specifically, the online degradation monitoring and triggering logic for the efficiency curves of the fuel cell stack, energy storage battery pack, and supercapacitor includes: sampling the average voltage and air equivalence ratio of the fuel cell stack at a frequency of 10Hz, triggering parameter identification when the accumulated stack charge reaches 3MWh. SOC (State of Charge) and internal resistance of the energy storage battery are sampled at a frequency of 5Hz, triggering EKF (Extended Kalman Filter) reassessment every 10,000km increase in accumulated mileage. ESR (Equivalent Series Resistance) of the supercapacitor is monitored at a frequency of 10Hz; if the peak discharge current reaches 85% of the rated value and occurs 1000 times, LMS (Adaptive Filter) updates are triggered.

[0026] For example, the process of automatically updating the efficiency curves of the fuel cell stack, energy storage battery pack, and supercapacitor after driving 10,000 km is as follows: When the vehicle's mileage reaches 10,116 km, the T-Box automatically triggers a "fixed-point calibration" task, capturing the fuel cell stack UI curves and battery pulse response data from a nearly 100 km stretch of uniform road. Recursive least squares and extended Kalman methods are used to re-estimate the polynomial coefficients. A 68kB update packet is generated and synchronized to the cloud via CAN-FD (Controller Area Network with Flexible Data-rate). If the overall net efficiency of the fuel cell stack decreases by more than 5% or the usable battery capacity decreases by more than 10%, the system issues a yellow alarm, prompting the fleet to schedule maintenance.

[0027] The updated results show that the net efficiency of the fuel cell stack decreased to 0.553 and 0.509 at 200kW and 300kW load points, respectively; the usable capacity of the energy storage battery pack decreased by 1.5%, and the internal resistance increased by 5mΩ, resulting in a decrease in the efficiency of the 200kW discharge channel to 0.923; the ESR of the supercapacitor increased by 12%, but its efficiency of the 200kW discharge channel decreased to 0.948.

[0028] After the update, and through 1000km real-vehicle verification, the fuel cell stack power consumption prediction error decreased from ±5.3% to ±2.1%. In C-WTVC (a driving cycle based on the transient cycle of heavy commercial vehicles worldwide, formed by adjusting acceleration / deceleration) simulation, the weighted hydrogen consumption per 100km decreased by approximately 0.12kg. Simultaneously, the system can issue a maintenance warning before the stack efficiency drops below 50% at 30,000km. This demonstrates that the online model update every 10,000km effectively ensures the consistency between the powertrain component model and actual performance, providing an accurate foundation for subsequent capacity rematching and topology switching strategies.

[0029] Step 2: Collect the real-time power demand of the vehicle, determine the typical operating conditions of the vehicle based on the real-time power demand, and then splice the power demand subsequences under each typical operating condition to obtain the vehicle power demand time series. The vehicle power demand time series serves as the constraint input for capacity matching and energy flow reconstruction.

[0030] Please see Figure 3 , Figure 3 The diagram shows a flowchart of the process for obtaining vehicle power demand time series data. Figure 3 As shown, in this embodiment, obtaining the vehicle power demand time series specifically includes the following steps: Step 2.1: Based on the vehicle's historical operating data, cluster analysis is performed to identify several typical operating conditions.

[0031] In this embodiment, typical operating conditions include: urban congestion, high-speed cruising, hill climbing with heavy load, long-duration cruising, and regenerative braking. The power demand data for each typical operating condition is described through feature extraction.

[0032] Specifically, in urban congestion conditions: power demand fluctuates greatly due to frequent start-stop cycles, ranging from approximately 0 to 120 kW. In highway cruising conditions: power demand is stable, primarily between 80 and 200 kW. In hill climbing and heavy-load conditions: power demand is high, between 260 and 300 kW, with a regenerative braking phase. In long-duration cruising conditions: the vehicle maintains a relatively long period of stable driving. In regenerative braking conditions: when the vehicle decelerates or brakes, power demand is negative, used for battery charging.

[0033] Step 2.2: Collect the real-time power demand of the vehicle and determine the typical operating conditions corresponding to the vehicle based on the real-time power demand.

[0034] In this embodiment, the real-time power demand of the vehicle is collected based on the actual route characteristics of the vehicle (such as mountain routes, intercity routes, or port routes). Based on the real-time power demand, relevant operating conditions are selected from the above five typical operating conditions to determine the typical operating conditions corresponding to the vehicle.

[0035] Step 2.3: Generate power demand subsequences for each typical operating condition by weighted fusion of data collected by the vehicle's on-board data recorder and bench simulation data.

[0036] In this embodiment, the vehicle's OBD (On-Board Data Recorder) collects vehicle speed, torque, and motor input power in real time at a frequency of 10Hz. To avoid excessive data volume and computational burden, the data is downsampled to 1Hz during post-processing, ensuring data accuracy while reducing computational resource requirements. Based on the actual vehicle operating data, simulation prediction models are used for data supplementation and prediction. The simulation data is based on different operating conditions to ensure the comprehensiveness and diversity of the data.

[0037] Understandably, to ensure the accuracy and comprehensiveness of the power demand sequence, this embodiment uses multi-source data fusion, combining real vehicle data and bench simulation data to form a unified power demand subsequence. By weighted fusion of real vehicle data and simulation data, the bias that may be caused by a single data source is eliminated, thereby more accurately simulating power demands under different operating conditions.

[0038] In this embodiment, a weighted average method is used to fuse data from different data sources. For data collected from actual vehicles, real-world operating conditions are prioritized; while simulation data is used to supplement operating conditions that cannot be covered by actual vehicle data, reflecting performance changes under extreme environments or after long-term operation. By optimizing the generated power demand time series, it is ensured that the series can serve as input for subsequent steps, guiding the design of system capacity matching and topology switching schemes. Based on this, it is guaranteed that the matched capacity parameters can adapt to different operating conditions such as urban start-stop efficiency, high-speed cruising economy, and heavy-load climbing dynamics.

[0039] Based on different typical operating conditions, the collected data was divided into three main stages. For example, the urban congestion condition: Under this condition, vehicles frequently encounter traffic lights and low-speed following situations, causing fluctuations in power demand. Data collection results show that the instantaneous drive power ranges from 0 to 120 kW, with an average power of approximately 45 kW. In this condition, the vehicle's power demand exhibits strong pulsation, with peak power occasionally reaching 110–120 kW during traffic light starts and stops and low-speed following. Continuous characteristics include 1600 sampling points, including approximately 40 complete stops, with an average dwell time of 22 seconds.

[0040] High-speed cruising condition: Under high-speed cruising conditions, the vehicle maintains a stable driving state, and the power demand is relatively stable. Data collected shows that the instantaneous power range is between 80–200kW, with an average power of approximately 140kW. During acceleration and overtaking, the power demand briefly increases to 220kW. The standard deviation of power fluctuation under this condition is approximately 7.5kW. The continuous characteristics include 1400 sampling points, with vehicle speed remaining stable between 78–82km / h and exhibiting relatively small variations.

[0041] Long Slope Heavy Load Condition: The vehicle carries a 45-ton load, requiring significant power during the climbing phase. Field test data shows that instantaneous power remained between 260–300kW, with an average power of 275kW. After climbing, the vehicle entered the regenerative braking phase, with regenerative braking power ranging from 40–60kW, averaging 45kW. The regenerative braking phase lasted 320 seconds, during which 320 power feedback points were collected.

[0042] Step 2.4: Concatenate the power demand subsequences according to the time sequence logic of the actual vehicle driving to obtain the vehicle power demand time series.

[0043] For example, the power data from the three different typical operating conditions mentioned above are spliced ​​together in chronological order to obtain a vehicle power demand time series with a total duration of 3500 seconds. This vehicle power demand time series contains 3500 power demand points and has a sampling resolution of 1Hz, which has high temporal resolution and can accurately reflect the fluctuation of vehicle power demand under different operating conditions.

[0044] Step 3: Use a mixed integer two-stage optimization algorithm to collaboratively optimize the capacity parameters and topology switching strategy of the multi-mode power system. The first stage of collaborative optimization uses the vehicle power demand time series as the constraint boundary to determine the capacity parameters. The second stage uses the switching state of the reconfigurable DC bus converter of the multi-mode power system as the integer decision variable to generate a topology switching calendar that matches the typical operating conditions corresponding to the vehicle power demand time series.

[0045] The mixed-integer two-stage optimization algorithm is a powerful mathematical tool widely used in problems requiring simultaneous optimization of continuous and discrete variables. In this embodiment, the optimization process of the mixed-integer two-stage optimization algorithm is divided into two stages: the first stage determines the capacity parameter, and the second stage determines the operating mode and topology. In this embodiment, the system's collaborative optimization process is described in [reference needed]. Figure 4 , Figure 4 The diagram shows a schematic of the system's collaborative optimization process.

[0046] In the first stage, the main focus is on optimizing the system's capacity parameters. In this embodiment, the capacity parameters include the rated power of the fuel cell stack, the capacity of the energy storage battery pack, and the capacity of the supercapacitor.

[0047] Specifically, the rated power of a fuel cell stack is the maximum power it can provide under rated operating conditions, directly affecting the system's power output capability. The capacity of the energy storage battery pack determines how much energy it can store, affecting the system's range and energy utilization efficiency under low-load conditions. The capacity of the supercapacitor determines the instantaneous power it can provide. It can provide a large amount of power in a short time, helping the system cope with instantaneous high power demands.

[0048] In the first phase, parameter optimization aims to maximize efficiency under low-load conditions. This involves adjusting the capacity configuration of system components when the vehicle load is light to achieve the lowest energy consumption and optimal system performance. Simultaneously, the overall vehicle weight and cost are considered, striving to minimize system weight and cost without sacrificing performance. For example, a larger battery pack can provide more energy but will increase weight and cost.

[0049] In the second stage, the optimization algorithm not only considers matching system capacity but also further improves system efficiency by adjusting operating modes and topology. By optimizing the topology, the system can dynamically switch operating modes according to actual power demand, ensuring efficient energy utilization under different operating conditions.

[0050] In this embodiment, one of the key objectives of the optimization algorithm is to maximize efficiency under low-load conditions. Low-load conditions refer to the vehicle's operating state when high power output is not required, such as low-speed driving on urban roads or long-term cruising. In these conditions, the fuel cell stack's efficiency is often low, and if the capacity and topology configuration are not appropriate, the system's energy efficiency and economy will be significantly reduced. Therefore, the optimization objective of this embodiment is to improve system efficiency under low-load conditions, ensuring that the power system maintains high efficiency while the vehicle is in motion, especially when power demand is low, thereby reducing energy waste. It also aims to reduce overall vehicle weight and cost by using appropriate capacity matching to avoid unnecessary weight increases caused by excessive battery capacity or other components. System costs will also be controlled during the optimization process, ensuring that the system is not only highly efficient but also competitive in the market.

[0051] It's important to note that the optimization process involves not only focusing on the system's energy efficiency but also comprehensively considering economic factors, such as minimizing mass and controlling costs. An excessively heavy powertrain leads to higher energy consumption; therefore, the optimization process strives to reduce the system's added mass. The optimization algorithm must also ensure that system costs remain within a reasonable range, avoiding unnecessary expenses due to excessive component additions. Through comprehensive consideration, the optimization algorithm can output a set of optimal system capacity parameters, ensuring excellent vehicle performance under various operating conditions while controlling mass and cost, achieving optimal economy and reliability.

[0052] In this embodiment, the mixed integer two-stage optimization algorithm uses a multi-objective function J as the optimization objective. The expression for the multi-objective function J is: ; in, Indicates time Vehicle power requirements, in units , For system efficiency, dimensionless. The time step is in seconds. Indicates the mass of the fuel cell stack. Indicates the mass of the energy storage battery pack. The mass of a supercapacitor is expressed in kilograms (kg). This indicates the power density of the fuel cell stack. This indicates the power density of the energy storage battery pack. The power density of a supercapacitor is expressed in units of... , Indicates the capacity of the fuel cell stack. Indicates the capacity of the energy storage battery pack. The capacitance of a supercapacitor is expressed in kilograms. This indicates the energy density of the fuel cell stack. This indicates the energy density of the energy storage battery pack. The energy density of a supercapacitor is expressed in units of 1000 kJ / m². , The first optimized weight coefficient, For the second optimization weighting coefficient, For the third optimization weight coefficient, Indicates time, This represents the total number of time steps.

[0053] In this embodiment, the first optimized weight coefficient Represents energy efficiency, second optimization weighting coefficient Representing system cost and quality, the third optimization weight coefficient This represents energy density. Among these factors, energy efficiency has the greatest impact on the system and carries a high weight, while system cost and quality have a secondary impact on overall optimization, and energy density is less important, although it can be crucial in certain application scenarios.

[0054] For example, the initial judgment matrix is ​​set as follows: The initial judgment matrix is ​​decomposed into eigenvalues, and the resulting eigenvectors are [0.5, 0.3, 0.2]. =0.5, =0.3, =0.2.

[0055] For actual system parameters, fuel cell stack: mass =450kg, power density =2.5kW / kg, capacity =120kg, energy density =500Wh / kg. Energy storage battery pack: mass =780kg, power density =3.0kW / kg, capacity =250kg, energy density =180Wh / kg. Supercapacitor: mass =165kg, power density =5.0kW / kg, capacity =70kg, energy density =25Wh / kg. System efficiency. =0.9, power demand, at each moment =100 W, time step Δt=1 s.

[0056] Substitute the above system parameters into the multi-objective function J for calculation:

[0057] If the first term of the multi-objective function J is T=1, then:

[0058] The second term of the multi-objective function J, the quality and cost part:

[0059] The third term of the multi-objective function J, the energy density part:

[0060] It should be noted that in the mixed-integer two-stage optimization algorithm, the results are combined with the specific constraints of system components for optimization calculation. By solving for the optimal solution through the optimization algorithm, a balance between energy efficiency, system mass, cost, and range under different topologies is obtained.

[0061] It is worth noting that in the optimization process of the multi-objective function J, upper limits are set for the health state of the energy storage battery and the voltage decay of the fuel cell. When either upper limit constraint is triggered, the weight coefficient of the second optimization is increased. This prioritizes reducing system mass and peak power output, thereby slowing down component aging rates.

[0062] In this embodiment, the second stage of the optimization process specifically includes the following steps: S1: Based on the determined capacity parameters, the energy flow topology state of the multi-mode power system is selected by taking the switching state of the reconfigurable DC bus converter of the multi-mode power system as an integer decision variable.

[0063] In this embodiment, the energy flow topology states are defined as follows: parallel connection of fuel cell stack and energy storage battery pack, direct drive of fuel cell stack and segmented connection of energy storage battery pack and supercapacitor.

[0064] Specifically, a direct-drive fuel cell stack is denoted as ID-1, meaning it is powered solely by the fuel cell. A fuel cell stack connected in parallel with a battery storage unit is denoted as ID-2, meaning both the fuel cell stack and the battery storage unit are powered together. A segmented battery storage unit and supercapacitor are denoted as ID-3, meaning the battery storage unit and supercapacitor switch discharges, and the fuel cell stack is idle.

[0065] S2: Jointly model the energy flow topology state with the continuous power allocation variables, and calculate the operating parameters under each energy flow topology state. The operating parameters include the power threshold and the energy storage battery state of charge window.

[0066] S3: Determine the continuous time periods that meet the operating parameters segment by segment along the time axis of the vehicle power demand time series, and generate a topology switching calendar based on the energy flow topology state and start and end times associated with each continuous time period.

[0067] S4: Calculate the switching cost corresponding to the topology switching calendar based on the topology switching penalty function, and output the topology switching calendar with the goal of minimizing the switching cost.

[0068] In this embodiment, the expression for the topology switching penalty function is: ; in, This represents the number of topology switches within the topology switch calendar period. This represents the energy loss during the i-th switching process. Let be the power interruption time caused by the i-th switching. This represents the average drive power within the 30 seconds prior to the switch.

[0069] For example, based on the determined capacity parameters: 300kW fuel cell stack, 120kWh energy storage battery pack, 5MJ supercapacitor, and a typical operating cycle of 3500s: 1600s of urban congestion + 1400s of highway cruising + 500s of long slope heavy load, with a sampling interval of 1s and a power demand range of 20~300kW, the process of obtaining its topology switching calendar is explained.

[0070] The optimization objectives and weights are set as follows: minimizing hydrogen consumption (weight 55%), minimizing vehicle added mass (weight 30%), minimizing system cost (weight 15%), and topology switching penalty coefficient (0.000035).

[0071] The lifespan and health status constraints are as follows: the state of charge (SOC) of the energy storage battery shall not be lower than 35% or higher than 80%, the energy storage battery health shall be maintained at ≥80%, and the average voltage decay of a single fuel cell shall not exceed 5%.

[0072] The in-loop strategy for a typical operating condition is shown below: In congested urban areas, when the instantaneous power is ≤80kW and the SOC is ≥60%, ID-3 is selected, with an average efficiency of ≈94.8%. In the 80~220kW range, switch to ID-2, maintaining the SOC at 60%–78%. When the peak power suddenly increases to >220kW, switch to ID-1 until the SOC drops to 55% and then return to ID-2.

[0073] During high-speed cruising from 1600 to 3000 seconds, the ID-2 level is maintained most of the time, and the SOC is automatically locked at 42%–65%.

[0074] During the long, heavily loaded section (3000-3500s), ID-1 is used continuously for the first 180s (260-300kW). During the downhill recovery phase (320s), ID-3 is switched to, recovering approximately 1.2MJ of energy.

[0075] A total of 5 topology switches occurred during the entire cycle, which is less than the limit of 3 times / hour. The average energy loss per switch was 0.28 MJ, the average downtime was 0.18 s, and the average power in the first 30 seconds of a switch was approximately 145 kW. The total weighted penalty contribution was less than 2% of the total target.

[0076] The optimization results are as follows: the equivalent hydrogen consumption per cycle is 7.94 kg / 100 km, which is 6.3% lower than the fixed parallel strategy. The state of charge (SOC) is maintained in the range of 36%–79%, the state of oxygen (SOH) of the energy storage battery decreases by 0.12%, and the average voltage of the fuel cell decreases by 0.04%. The calendar switching successfully takes into account all constraints, ensuring low-load efficiency and overall vehicle economy.

[0077] The output topology switching calendar is recorded in the format of "start time - end time - topology ID": 0~460s: ID~3; 460~1110s: ID~2; 1110~1160s: ID~1; 1160~1600s: ID~2; 1600~3000s: ID~2; 3000~3180s: ID~1; 3180~3500s: ID~3; The topology switching calendar is deployed to the vehicle controller and synchronized to the fleet cloud via OTA, enabling precise topology switching management for complex operating conditions.

[0078] Step 4: Input the capacity parameters and topology switching calendar into the vehicle simulation test platform for performance verification. If the simulation results do not meet the preset target, adjust the optimization weight coefficients of the mixed integer two-stage optimization algorithm and re-execute step 3 until the simulation results meet the preset target.

[0079] In this embodiment, the vehicle simulation test platform can use the hardware-in-the-loop Monte Carlo joint verification method to perform 1000 batch simulations on random combination conditions. The random combination conditions are ambient temperature -20℃~40℃, road slope 0%~6% and vehicle load 70%~100%.

[0080] In this embodiment, the preset targets include hydrogen consumption per 100 kilometers and acceleration time. For example, the preset targets are hydrogen consumption per 100 kilometers ≤ 8.5 kg and acceleration time from 0-80 km / h ≤ 35 s.

[0081] The parameter matching method for the multi-mode power system of hydrogen-powered heavy-duty trucks in this embodiment uses a "capacity-topology" integrated matching mechanism to collaboratively determine the rated power of the fuel cell stack, the capacity of the power battery pack, and the capacity of the supercapacitor under the same mixed integer two-stage optimization framework. Combined with the topology adaptive reconfiguration of the reconfigurable DC bus converter, it can significantly reduce energy loss and redundant mass under low-load conditions while meeting peak power demand. Compared with the existing schemes that rely on experience-based selection, it can reduce the vehicle's added weight by about 10%-15%, reduce the overall component cost by about 8%, and improve the average system efficiency to over 53% in urban congestion cycles. It fundamentally solves the problem of efficiency decline and excessive cost caused by the lack of a systematic method for matching the capacity of fuel cell stacks and energy storage devices.

[0082] The parameter matching method for the multi-mode power system of hydrogen-powered heavy-duty trucks in this embodiment introduces innovative designs such as online model self-correction based on efficiency curve decay, topology switching penalty function, and dual constraints on the lifespan of energy storage battery and fuel cell. This achieves adaptive adjustment of the "power threshold - energy storage battery SOC window" according to operating conditions and health status. With hardware-in-the-loop - Monte Carlo joint verification, it can stably maintain hydrogen consumption ≤8.5kg per 100km and 0-80km / h acceleration ≤35s under random combinations of operating conditions of -20℃~40℃, 0%~6% gradient, and 70%~100% load. It significantly suppresses energy consumption fluctuations under frequent start-stop and complex road conditions, and slows down the aging rate of core components, comprehensively improving the economy, reliability, and adaptability of the multi-mode power system of hydrogen-powered heavy-duty trucks.

[0083] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations are intended to cover non-exclusive inclusion, such that an article or device comprising a list of elements includes not only those elements but also other elements not expressly listed. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the article or device comprising said element. Terms such as "connected" or "linked" are not limited to physical or mechanical connections but can include electrical connections, whether direct or indirect. The orientations or positional relationships indicated by terms such as "upper," "lower," "left," and "right" are based on the orientations or positional relationships shown in the accompanying drawings and are used only for the convenience of describing the invention and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as limiting the invention.

[0084] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features or characteristics described may be combined in any suitable manner in one or more embodiments or examples. In addition, those skilled in the art can combine and integrate the different embodiments or examples described in this specification.

[0085] Those skilled in the art will recognize that the embodiments described herein are intended to help the reader understand the principles of the invention, and should be understood that the scope of protection of the invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific modifications and combinations based on the technical teachings disclosed in this invention without departing from the spirit of the invention, and these modifications and combinations are still within the scope of protection of this invention.

Claims

1. A method for parameter matching of a multi-mode power system for hydrogen-powered heavy-duty trucks, characterized in that, include: Step 1: Construct power system component models for the multi-mode power system of hydrogen-powered heavy trucks and set up online model self-calibration strategies; Step 2: Collect the real-time power demand of the vehicle, determine the typical operating conditions of the vehicle based on the real-time power demand, and then splice the power demand subsequences under each typical operating condition to obtain the vehicle power demand time series. The vehicle power demand time series serves as the constraint input for capacity matching and energy flow reconstruction. Step 3: Use a mixed integer two-stage optimization algorithm to perform collaborative optimization of the capacity parameters and topology switching strategy of the multi-mode power system. The first stage of the collaborative optimization determines the capacity parameters with the vehicle power demand time series as the constraint boundary. The second stage uses the switching state of the reconfigurable DC bus converter of the multi-mode power system as the integer decision variable to generate a topology switching calendar that matches the typical operating conditions corresponding to the vehicle power demand time series. Step 4: Input the capacity parameters and the topology switching calendar into the vehicle simulation test platform for performance verification. If the simulation results do not meet the preset target, adjust the optimization weight coefficients of the mixed integer two-stage optimization algorithm and re-execute step 3 until the simulation results meet the preset target.

2. The parameter matching method for a multi-mode power system of a hydrogen-powered heavy-duty truck according to claim 1, characterized in that, The power system component model includes a fuel cell stack, an energy storage battery pack, a supercapacitor, and a reconfigurable DC bus converter.

3. The parameter matching method for a multi-mode power system of a hydrogen-powered heavy-duty truck according to claim 2, characterized in that, The online model self-calibration strategy includes automatically updating the efficiency curves of the fuel cell stack, the energy storage battery pack, and the supercapacitor every 10,000 km of cumulative vehicle mileage to compensate for model errors caused by component performance degradation.

4. The parameter matching method for a multi-mode power system of a hydrogen-powered heavy-duty truck according to claim 1, characterized in that, Step 2 includes: Step 2.1: Based on historical vehicle operation data, cluster analysis is performed to identify several typical operating conditions; Step 2.2: Collect the real-time power demand of the vehicle, and determine the typical operating conditions corresponding to the vehicle based on the real-time power demand; Step 2.3: Generate power demand subsequences for each typical operating condition by weighted fusion of data collected by the on-board data recorder and bench simulation data; Step 2.4: Concatenate the power demand subsequences according to the time sequence logic of the actual vehicle driving to obtain the vehicle power demand time sequence.

5. The parameter matching method for a multi-mode power system of a hydrogen-powered heavy-duty truck according to claim 1, characterized in that, The capacity parameters include the rated power of the fuel cell stack, the capacity of the energy storage battery pack, and the capacity of the supercapacitor.

6. The parameter matching method for a multi-mode power system of a hydrogen-powered heavy-duty truck according to claim 1, characterized in that, The mixed-integer two-stage optimization algorithm uses a multi-objective function J as the optimization objective, and the expression of the multi-objective function J is: ； in, Indicates time Vehicle power requirements For system efficiency, For time step, Indicates the mass of the fuel cell stack. Indicates the mass of the energy storage battery pack. Indicates the mass of a supercapacitor. This indicates the power density of the fuel cell stack. This indicates the power density of the energy storage battery pack. This indicates the power density of a supercapacitor. Indicates the capacity of the fuel cell stack. Indicates the capacity of the energy storage battery pack. This indicates the capacitance of the supercapacitor. This indicates the energy density of the fuel cell stack. This indicates the energy density of the energy storage battery pack. This indicates the energy density of a supercapacitor. The first optimized weight coefficient, For the second optimization weighting coefficient, For the third optimization weight coefficient, Indicates time, This represents the total number of time steps.

7. The parameter matching method for a multi-mode power system of a hydrogen-powered heavy-duty truck according to claim 6, characterized in that, In the optimization process of the multi-objective function J, upper limits of the health status of the energy storage battery and the upper limits of the voltage decay of the fuel cell are set. When either upper limit constraint is triggered, the second optimization weight coefficient is increased.

8. The parameter matching method for a multi-mode power system of a hydrogen-powered heavy-duty truck according to claim 1, characterized in that, Using the switching state of the reconfigurable DC bus converter of the multi-mode power system as an integer decision variable, a topology switching calendar matching the typical operating conditions corresponding to the vehicle power demand time series is generated, including: Based on the determined capacity parameters, the energy flow topology state of the multi-mode power system is selected by taking the switching state of the reconfigurable DC bus converter of the multi-mode power system as an integer decision variable. The energy flow topology state is jointly modeled with the continuous power allocation variable, and the operating parameters under each energy flow topology state are calculated. The operating parameters include the power threshold and the energy storage battery state of charge window. The continuous time periods that meet the operating parameters are determined segment by segment along the time axis of the vehicle power demand time series. The topology switching calendar is generated based on the energy flow topology state and start and end times associated with each continuous time period. The switching cost corresponding to the topology switching calendar is calculated based on the topology switching penalty function, and the topology switching calendar is output with the goal of minimizing the switching cost.

9. The parameter matching method for a multi-mode power system of a hydrogen-powered heavy-duty truck according to claim 8, characterized in that, The expression for the topology switching penalty function is: ； in, This represents the number of topology switches within the topology switch calendar period. This represents the energy loss during the i-th switching process. Let be the power interruption time caused by the i-th switching. This represents the average drive power within the 30 seconds prior to the switch.

10. The parameter matching method for a multi-mode power system of a hydrogen-powered heavy-duty truck according to claim 1, characterized in that, The preset targets include hydrogen consumption per 100 kilometers and acceleration time.

Images