Parameter optimization method for energy management system of fuel cell and lithium battery hybrid power

Abstract

The invention relates to the technical field of lithium batteries, in particular to a fuel battery and lithium battery hybrid power energy management system parameter optimization method, which comprises the following steps: acquiring working state data and load demand data of a hybrid power system; acquiring real-time performance parameters of the fuel cell and the lithium battery; determining a system operation mode based on the working state data and the load demand data; executing a corresponding control strategy according to the system operation mode; dynamically adjusting parameters of a control strategy based on the real-time performance parameters; generating an optimized power distribution instruction; and sending the power distribution instruction to an execution unit of the hybrid power system. According to the method provided by the invention, an optimization strategy based on dynamic planning and model prediction control is introduced, so that the calculation efficiency is greatly improved while the optimization effect is ensured, and real-time optimization becomes possible.

Description

Technical Field

[0001] The present invention relates to the technical field of lithium batteries, and more specifically, to a method for optimizing the parameters of an energy management system for a fuel cell lithium battery hybrid power system. Background Art

[0002] With the improvement of environmental awareness and the development of renewable energy technologies, fuel cell lithium battery hybrid power systems have been increasingly valued in various application fields. This hybrid power system combines the high energy density of fuel cells and the high power density characteristics of lithium batteries, and theoretically can achieve a perfect balance between long endurance and high performance. However, how to effectively manage the collaborative work of these two energy sources and how to optimize system parameters to achieve the best performance have always been the research hotspots and challenges in this field.

[0003] Existing energy management systems usually adopt rule-based control strategies or simple optimization algorithms. These methods can achieve good results under specific working conditions, but often seem powerless in the face of complex and changing actual operating environments. For example, rule-based control strategies are difficult to adapt to rapid changes in load and long-term drift of system parameters; while simple optimization algorithms often can only consider a single objective and are difficult to achieve a balance in multiple aspects such as fuel economy, system efficiency, and lifespan.

[0004] In addition, the existing methods generally have the following problems: First, the consideration of system dynamic characteristics is insufficient, resulting in poor performance under transient working conditions; second, there is a lack of effective adaptive mechanisms, making it difficult to cope with changes in system parameters and external environmental disturbances; third, the attention to the long-term performance of the system is insufficient, often only pursuing short-term benefits while ignoring long-term indicators such as system lifespan; finally, it is difficult to achieve a good balance between computational efficiency and optimization effect in the existing methods, either the real-time performance is insufficient, or the optimization results are not ideal enough.

[0005] In view of the above problems, there is an urgent need for an energy management system parameter optimization method that can comprehensively consider system characteristics, has an adaptive ability, takes into account both short-term and long-term performance, and is computationally efficient. The present invention is precisely proposed in response to this need. Summary of the Invention

[0006] The method for optimizing the parameters of an energy management system for a fuel cell lithium battery hybrid power system proposed by the present invention effectively solves the problems existing in the prior art by comprehensively applying advanced control theories and optimization algorithms. This method can not only achieve dynamic optimization of system parameters but also make adaptive adjustments according to real-time working conditions, thus significantly improving the overall performance of the hybrid power system.

[0007] The present invention provides a method for optimizing the parameters of an energy management system for a fuel cell lithium battery hybrid power system, including:

[0008] The acquisition step includes:

[0009] Obtain the working state data and load demand data of the hybrid power system;

[0010] Obtain the real-time performance parameters of the fuel cell and the lithium battery;

[0011] The processing step includes:

[0012] Based on the working state data and the load demand data, determine the system operation mode;

[0013] According to the system operation mode, execute the corresponding control strategy;

[0014] Based on the real-time performance parameters, dynamically adjust the parameters of the control strategy;

[0015] The output step includes:

[0016] Generate an optimized power distribution instruction;

[0017] Send the power distribution instruction to the execution unit of the hybrid power system.

[0018] Preferably, the control strategy includes at least one of a PID control strategy, a state machine decoupling control strategy, an ECMS equivalent hydrogen minimum consumption strategy, and an optimization control strategy based on DP-MPC.

[0019] Preferably, the PID control strategy specifically includes:

[0020] Obtain the error signal between the system output power and the target power;

[0021] Based on the error signal, calculate the proportional term, the integral term, and the differential term;

[0022] Generate a control signal according to the weighted sum of the proportional term, the integral term, and the differential term.

[0023] Preferably, the state machine decoupling control strategy specifically includes:

[0024] Based on the working state data, determine the current state of the system;

[0025] According to the state, select a preset power distribution strategy;

[0026] Based on the preset power distribution strategy, generate the power distribution ratio of the fuel cell and the lithium battery.

[0027] Preferably, the ECMS equivalent hydrogen minimum consumption strategy specifically includes:

[0028] Obtain the instantaneous hydrogen consumption rate of the fuel cell and the equivalent hydrogen consumption rate of the lithium battery;

[0029] Based on the instantaneous hydrogen consumption rate and the equivalent hydrogen consumption rate, an optimization objective function is constructed;

[0030] Solve the optimization objective function to obtain an optimal power distribution scheme.

[0031] Preferably, the optimization control strategy based on DP-MPC specifically includes:

[0032] Based on the current system state, predict the load demand for a period of time in the future;

[0033] Use the dynamic programming algorithm to solve the optimal control sequence within the prediction time domain;

[0034] Execute the first control quantity of the optimal control sequence and repeat the optimization process in the next control cycle.

[0035] Preferably, it further includes a parameter optimization step:

[0036] Obtain the historical data of system operation;

[0037] Based on the historical data, construct a system performance evaluation index;

[0038] Use a multi-objective optimization algorithm to optimize the key parameters of the control strategy.

[0039] Preferably, the system performance evaluation index includes at least one of hydrogen consumption, energy utilization rate, power stability, and system life.

[0040] Preferably, it further includes an adaptive parameter adjustment step:

[0041] Real-time monitor the system operation state and performance indicators;

[0042] Based on the change trends of the operation state and performance indicators, dynamically adjust the parameters of the control strategy;

[0043] Apply the adjusted parameters to the next control cycle.

[0044] Preferably, it further includes a model update step:

[0045] Collect the actual operation data of the system;

[0046] Based on the actual operation data, use an online parameter identification method to update the system model;

[0047] Use the updated system model for subsequent optimization of the control strategy.

[0048] The beneficial effects of the present invention are mainly reflected in the following aspects:

[0049] The method of the present invention builds an accurate system model, fully considers the dynamic characteristics of fuel cells and lithium batteries, and can effectively respond to rapid changes in the load. By introducing a multi-objective optimization algorithm, this method achieves a balance in multiple aspects such as fuel economy, system efficiency, power stability, and lifespan. In particular, the method of the present invention introduces an optimization strategy based on dynamic programming and model predictive control, which not only ensures the optimization effect but also greatly improves the computational efficiency, making real-time optimization possible.

[0050] In addition, the method of the present invention also has significant adaptive capabilities. By real-time monitoring the system state and updating the model parameters, this method can effectively respond to long-term changes in system performance and external environmental disturbances. This adaptive mechanism not only improves the robustness of the system but also extends the service life of the entire hybrid power system.

[0051] It is worth mentioning that while achieving the above optimization goals, the method of the present invention also maintains a low computational complexity. This feature enables this method to run in real-time on an embedded system, greatly expanding its application scope.

[0052] In summary, the energy management system parameter optimization method proposed by the present invention effectively solves the problems existing in the prior art through a number of innovative designs. This method not only improves the energy utilization efficiency of the hybrid power system but also significantly enhances the adaptability and reliability of the system. These improvements will provide strong technical support for the popularization and application of fuel cell lithium battery hybrid power systems in various application scenarios, and have important theoretical significance and practical application value. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 is a flowchart of the method of the present invention.

[0054] Figure 2 is a logic diagram of the control strategy selection in the processing steps of the present invention.

[0055] Figure 3 is a logic block diagram of the optimization steps of the present invention.

[0056] Figure 4 is a logic block diagram of the adaptive parameter adjustment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0057] In order to further elaborate on the technical means and effects adopted by the present invention to achieve the intended invention purpose, the following combines the accompanying drawings and preferred embodiments to detail its specific implementation manner, structure, features, and effects as follows. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0058] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the technical field to which this invention pertains.

[0059] Please refer to Figures 1-4 , the present invention discloses a method for optimizing the parameters of an energy management system for a fuel cell - lithium battery hybrid power system. The method aims to improve the energy efficiency of the hybrid power system, extend the system life, and achieve dynamic optimization control.

[0060] Specifically, the method of the present invention includes an acquisition step, a processing step, and an output step. In the acquisition step, first, the operating state data and load demand data of the hybrid power system are acquired. These data can be collected in real - time through various sensors, such as voltage sensors, current sensors, temperature sensors, etc. Preferably, the operating state data includes parameters such as the output voltage, output current, and temperature of the fuel cell and the lithium battery. The load demand data reflects the current power demand of the system and can be obtained through a load power sensor.

[0061] Next, the method of the present invention also acquires the real - time performance parameters of the fuel cell and the lithium battery. These parameters are crucial for evaluating the health state and performance of the battery. For the fuel cell, parameters such as its membrane conductivity and catalyst activity can be acquired; for the lithium battery, parameters such as its internal resistance and remaining capacity can be acquired. These performance parameters can be obtained through a dedicated battery management system (BMS).

[0062] In the processing step, the method of the present invention first determines the operating mode of the system based on the acquired operating state data and load demand data. For example, the system may be in a start - up mode, a normal operating mode, a high - power demand mode, or a low - power demand mode, etc. The process of determining the operating mode can be achieved through preset rules or machine - learning algorithms. For example, if the load demand suddenly increases to more than 90% of the system rated power, it can be determined that the system enters the high - power demand mode.

[0063] Then, according to the determined operating mode of the system, the method of the present invention executes the corresponding control strategy. The present invention provides a variety of control strategies, including PID control strategy, state - machine decoupling control strategy, ECMS equivalent hydrogen minimum consumption strategy, and DP - MPC - based optimization control strategy. The selection of these strategies depends on the operating mode and performance requirements of the system. For example, in the normal operating mode, the ECMS strategy may be preferably selected to minimize hydrogen consumption; while in the high - power demand mode, the DP - MPC - based optimization control strategy may be selected to quickly respond to load changes.

[0064] Furthermore, the method of the present invention dynamically adjusts control strategy parameters based on real-time performance parameters. This adaptive adjustment mechanism enables the system to better adapt to changes in battery performance and aging. For example, if a decrease in catalyst activity in a fuel cell is detected, its output power limit can be adjusted accordingly to avoid overuse and extend its lifespan.

[0065] In the output step, the method of the present invention generates optimized power allocation instructions. These instructions determine the power output of the fuel cell and the lithium battery, respectively. Preferably, the power allocation instructions are given as percentages, for example, the fuel cell bears 70% of the load and the lithium battery bears 30%. Finally, these power allocation instructions are sent to the hybrid system's execution unit, such as a DC / DC converter or power electronics interface, to implement actual power control.

[0066] The PID control strategy of the present invention is a classic feedback control method. Specifically, the strategy first obtains the error signal between the system output power and the target power. The error signal e(t) can be expressed as:

[0067] e(t)=P target (t)-P output (t),

[0068] The PID control strategy of the present invention is a classic feedback control method. Specifically, the strategy first obtains the error signal between the system output power and the target power. The error signal e(t) can be expressed as: target (t) is the target power, P output (t) is the actual output power.

[0069] Then, based on the error signal, the proportional term, integral term and differential term are calculated. These three terms are expressed as follows: Proportional term: P = K p e(t); integral term: Differential term: Among them, K p ,K i ,K d These are the proportional gain, integral gain, and derivative gain, respectively. The selection of these parameters is crucial to the performance of the PID controller. Generally, the optimal values for these parameters can be determined using the Ziegler-Nichols tuning method or other optimization algorithms.

[0070] Finally, the control signal u(t) is generated based on the weighted sum of these three terms:

[0071]

[0072] This control signal is used to adjust the output power of the fuel cell and the lithium battery, so that the actual output power of the system is close to the target power.

[0073] The PID control strategy of the present invention has the advantages of simple structure and strong robustness, and is applicable to various operating modes. Especially when the system load changes little, the PID control can provide stable and fast response. However, for the cases with high non-linearity or drastic load changes, other control strategies may need to be combined to obtain better performance.

[0074] The state machine decoupling control strategy of the present invention is a hierarchical control method based on the system state. This strategy first determines the current state of the system based on the working state data. Preferably, the present invention defines the following several states:

[0075] 1. Startup state: The system has just started, and the fuel cell has not reached the optimal operating temperature.

[0076] 2. Normal operating state: The system operates stably within the rated power range.

[0077] 3. High power demand state: The load power exceeds 80% of the system rated power.

[0078] 4. Low power demand state: The load power is lower than 20% of the system rated power.

[0079] 5. Emergency state: The system has a fault or anomaly.

[0080] The judgment of the state can be achieved by setting appropriate thresholds. For example, when it is detected that the load power suddenly increases and exceeds 80% of the rated power, the system enters the high power demand state. The selection of these thresholds needs to be adjusted according to the specific system characteristics and application scenarios.

[0081] After determining the current state, the method of the present invention will select a preset power distribution strategy. The power distribution strategy in each state is optimized to meet the specific requirements in that state. For example:

[0082] In the startup state, it mainly relies on the lithium battery for power supply, and at the same time gradually increases the output power of the fuel cell. This can be expressed as:

[0083]

[0084] P batt =P load -P fc

[0085] Here, P fc and P batt represent the output powers of the fuel cell and the lithium battery respectively; is the maximum output power of the fuel cell; T warmup is the preset preheating time; P load is the load power.

[0086] Normal operating state

[0087]

[0088] Under normal operating conditions, the fuel cell is preferentially used for power supply and maintained near its optimal operating point: where represents the optimal operating power point of the fuel cell, which is usually near the peak of its efficiency curve.

[0089] Under high power demand conditions, the strategy may become:

[0090]

[0091] This strategy ensures that the fuel cell operates at its maximum output power, while the lithium battery supplements the remaining power demand.

[0092] Based on the selected power distribution strategy, the method of the present invention generates the power distribution ratios of the fuel cell and the lithium battery. These ratios can be directly used to control the output power of each power source.

[0093] The advantage of the state machine decoupling control strategy is that its logic is clear, easy to implement and debug. By decomposing complex control problems into multiple relatively simple states, this strategy can effectively handle the requirements of the system under different operating conditions. However, this strategy also has some limitations. For example, discontinuous control outputs may occur during state transitions. Therefore, in practical applications, it is usually necessary to combine other control strategies, such as PID control or model predictive control, to obtain smoother and more optimized control effects. The state machine decoupling control strategy of the present invention further includes a dynamic adjustment function. During actual operation, the system may encounter situations that cannot be fully covered by the preset states. Therefore, the method of the present invention introduces a fuzzy processing mechanism for state boundaries. Specifically, when the system state approaches the boundary between two preset states, a weighted average method is used to calculate the power distribution ratio. This processing method can smooth the state transition process and avoid sudden changes in control outputs.

[0094] For example, assume that the current load power of the system is between the normal operating state and the high power demand state, and a transition function f(P load ):

[0095] [[ID=�8]]

[0096] where P threshold_low and P threshold_highThey are the low threshold and high threshold of state transition respectively. Then, the final power distribution can be expressed as:

[0097]

[0098] P batt = P load - P fc ,

[0099] where and are the output powers of the fuel cell in the normal operation state and the high power demand state respectively.

[0100] The ECMS equivalent hydrogen minimum consumption strategy of the present invention is an energy management method based on instantaneous optimization. The core idea of this strategy is to convert the electric energy consumption of the battery into equivalent hydrogen consumption, thus transforming the problem into an optimization problem with a single objective.

[0101] When implementing the ECMS strategy, the method of the present invention first obtains the instantaneous hydrogen consumption rate of the fuel cell and the equivalent hydrogen consumption rate of the lithium battery. The hydrogen consumption rate of the fuel cell can be calculated by the following formula:

[0102]

[0103] where m H2,fc is the hydrogen consumption rate of the fuel cell (g / s), P fc is the output power of the fuel cell (W), η fc is the efficiency of the fuel cell, and LHV H2 is the lower heating value of hydrogen (about 120 MJ / kg).

[0104] The equivalent hydrogen consumption rate of the lithium battery can be expressed as:

[0105]

[0106] where m H2,batt is the equivalent hydrogen consumption rate of the lithium battery, P batt is the output power of the lithium battery, η batt is the efficiency of the lithium battery, and s(t) is the equivalent factor. The equivalent factor s(t) is a key parameter of the ECMS strategy, which reflects the conversion relationship between battery energy and hydrogen energy. Based on the above consumption rates, the method of the present invention constructs an optimization objective function:

[0107] J = m H2,fc + m H2,batt ,

[0108] The optimization problem can be expressed as:

[0109]

[0110] Among them, P load is the load power, P fc,min and P fc,max are the minimum and maximum output power of the fuel cell, P b att, min and P batt,max They are the minimum and maximum output power of lithium batteries, SOC min and SOC max They are the minimum and maximum states of charge allowed for lithium batteries.

[0111] To solve this optimization problem, the method of the present invention adopts the golden section search algorithm, which can quickly find the local optimal solution and is suitable for real-time control requirements.

[0112] The output of the optimization process is the optimal power allocation solution. A key challenge of the ECMS strategy is the selection of the equivalent factor s(t). In a preferred embodiment of the present invention, an adaptive ECMS (A-ECMS) method is used to dynamically adjust the equivalent factor. Specifically, the equivalent factor is updated based on the deviation of the battery SOC:

[0113] s(t)=s0+K p (SOC ref -SOC)+K i ∫(SOC ref -SOC)dt,

[0114] Where s0 is the initial equivalent factor, SOC ref is the target SOC value, K p and K i are proportional and integral gains respectively. This adaptive mechanism enables the system to maintain the battery SOC within a reasonable range during long-term operation.

[0115] The DP-MPC-based optimization control strategy of the present invention combines the global optimality of dynamic programming (DP) with the real-time performance of model predictive control (MPC) and is an advanced energy management method. The implementation process of this strategy can be divided into the following steps:

[0116] First, based on the current system state, the load demand for a period of time in the future is predicted. The present invention uses an improved long short-term memory network (LSTM) for load forecasting. The input of the forecasting model includes historical load data, time information and possible environmental factors. The forecast step size N p It is usually chosen between 10 and 30 seconds, which can achieve a good balance between prediction accuracy and computational complexity.

[0117] Next, a dynamic programming algorithm is used to solve the optimal control sequence within the prediction horizon. The state variables of dynamic programming include the fuel cell output power P fc and the state of charge (SOC) of the battery. The control variable is the power change rate ΔP fc of the fuel cell. The objective function is defined as:

[0118]

[0119] where m H2 (k) is the hydrogen consumption at the k-th step, C deg (k) is the degradation cost of the fuel cell, SOC(k) is the state of charge of the battery, SOC ref is the target state of charge, and α, β, and γ are weight coefficients. The recurrence equation of dynamic programming can be expressed as:

[0120]

[0121] where x k is the system state, u k is the control input, and g(x k , u k ) is the stage cost function. After obtaining the optimal control sequence, the method of the present invention executes the first control quantity of the sequence. This rolling optimization method enables the system to respond promptly to changes in the actual working conditions. In the next control cycle, the system will update the state information, re-perform prediction and optimization, so as to achieve closed-loop control.

[0122] A key advantage of the DP-MPC strategy is its ability to consider both short-term and long-term performance goals simultaneously. By adjusting the weights of each item in the objective function, a good balance can be achieved among fuel economy, battery life, and dynamic response. For example, when the battery SOC is low, the value of γ can be increased to encourage the system to charge; when the fuel cell is severely aged, the value of β can be increased to reduce the use of the fuel cell.

[0123] The method of the present invention further includes a parameter optimization step, aiming to further improve the performance of the energy management system. This step first obtains the historical operation data of the system, including information such as load power, output power, efficiency, and temperature of the fuel cell and lithium battery. These data are usually sampled at a high frequency (such as 1 Hz) and preprocessed to remove outliers and noise.

[0124] Based on the obtained historical data, the method of the present invention constructs system performance evaluation indicators. These indicators comprehensively reflect the operation effect of the system and mainly include:

[0125] 1. Hydrogen consumption: reflecting the fuel economy of the system, the calculation formula is:

[0126]

[0127] 2. Energy utilization efficiency: Measures the energy utilization efficiency of the system and is defined as:

[0128]

[0129] 3. Power stability: Evaluates the degree of fluctuation of the system output power and can be represented by the standard deviation of power:

[0130]

[0131] 4. System life index: Considering the degradation of fuel cells and lithium batteries, it can be expressed as:

[0132] L sys =min(L fc ,L batt ),

[0133] where L fc and L batt are the estimated remaining lives of the fuel cell and the lithium battery respectively. After constructing these evaluation indicators, the method of the present invention uses a multi-objective optimization algorithm to optimize the key parameters of the control strategy. Preferably, an improved non-dominated sorting genetic algorithm (NSGA-II) is used for optimization. The objective function of this algorithm can be expressed as:

[0134] minF(x)=[f1(x),f2(x),f3(x),f4(x)],

[0135] where x is the parameter vector to be optimized, and f1, f2, f3, and f4 correspond to the above four evaluation indicators respectively.

[0136] The main steps of the NSGA-II algorithm include:

[0137] 1. Initialize the population;

[0138] 2. Non-dominated sorting;

[0139] 3. Crowding degree calculation;

[0140] 4. Selection;

[0141] 5. Crossover and mutation;

[0142] 6. Elite retention;

[0143] Through multiple iterations, the algorithm finally obtains a set of Pareto optimal solutions. The system administrator can select the most suitable parameter configuration from this set of solutions according to actual needs.

[0144] The parameter optimization steps of the present invention can significantly improve the overall performance of the energy management system. By comprehensively considering multiple objectives, this method can find the best balance among fuel economy, system efficiency, stability, and lifespan. In addition, since the optimization process is carried out offline, more complex algorithms and longer calculation times can be used, resulting in better results. When constructing the system performance evaluation index of the present invention, multiple aspects such as hydrogen consumption, energy utilization rate, power stability, and system lifespan are comprehensively considered. The selection of these indicators aims to comprehensively reflect the operation effect of the fuel cell - lithium battery hybrid system and provide a reliable basis for subsequent parameter optimization.

[0145] Specifically, hydrogen consumption is a direct indicator for evaluating the fuel economy of the system. In the preferred embodiment of the present invention, the calculation of hydrogen consumption not only considers the direct consumption of the fuel cell but also includes the equivalent consumption during the charge - discharge process of the lithium battery. This comprehensive calculation method can more accurately reflect the true energy consumption of the system.

[0146] The energy utilization rate reflects the energy conversion efficiency of the system. The present invention uses the ratio of output power to input power to characterize the energy utilization rate. This method is simple and intuitive, easy to calculate and monitor in real - time. In practical applications, the energy utilization rate usually fluctuates with the change of load power. Therefore, the method of the present invention calculates the average energy utilization rate within a certain time period (such as 15 minutes) to eliminate the influence of short - term fluctuations.

[0147] Power stability is an important indicator for measuring the dynamic performance of the system. The present invention uses the standard deviation of power output to quantify power stability. This method not only considers the amplitude of power fluctuations but also reflects the frequency of fluctuations. In actual operation, different time windows can be selected to calculate the standard deviation according to different application scenarios. For example, for applications requiring fast response, a shorter time window (such as 1 second) can be selected; while for applications emphasizing long - term stability, a longer time window (such as 1 minute) can be selected.

[0148] The system lifespan indicator reflects the attention of the present invention to the long - term reliability of the system. This indicator comprehensively considers the degradation of the fuel cell and the lithium battery. During the calculation process, the method of the present invention establishes a degradation model based on the usage conditions of each component (such as charge - discharge times, depth, temperature, etc.) to estimate the remaining lifespan. It should be noted that the system lifespan depends on the key component with the shortest lifespan. Therefore, the present invention uses the minimum value as the system lifespan indicator.

[0149] The selection and calculation methods of these performance evaluation indicators reflect the comprehensiveness and forward - looking nature of the present invention. Through these indicators, system managers can comprehensively understand the operation status of the hybrid system and provide strong support for subsequent optimization and maintenance.

[0150] The method of the present invention further includes an adaptive parameter adjustment step, which is the key to improving the adaptability and robustness of the system. This step first monitors the operating state and performance indicators of the system in real time. The monitored parameters include but are not limited to: the output voltage, current, and temperature of the fuel cell and lithium battery, the system output power, the load demand, and the various performance indicators mentioned above.

[0151] Based on these monitored data, the method of the present invention analyzes the change trends of the operating state and performance indicators. This process adopts the method of a sliding time window, taking into account both short-term fluctuations and long-term trends. For example, for the power stability indicator, the moving averages of 1 minute, 15 minutes, and 1 hour can be calculated simultaneously to comprehensively grasp the dynamic characteristics of the system.

[0152] Based on the analysis, the method of the present invention dynamically adjusts the parameters of the control strategy. This adjustment is continuous and smooth to avoid violent fluctuations in the system. Taking the PID controller as an example, its parameter adjustment can be expressed as:

[0153] K p (t) = K p0 + ΔK p (t);

[0154] K i (t) = K i0 + ΔK i (t);

[0155] K d (t) = K d0 + ΔK d (t);

[0156] Wherein, K p0 , K i0 , K d0 are the initial parameter values, and ΔK p (t), ΔK i (t), ΔK d (t) are the dynamic adjustment amounts. These adjustment amounts can be calculated through intelligent algorithms such as fuzzy logic or neural networks.

[0157] Preferably, the method of the present invention further introduces a constraint mechanism for parameter adjustment to ensure that the system always remains within a safe and stable operating range. For example, an upper limit of the parameter change rate can be set:

[0158]

[0159] Where ∈ p , ∈ i , ∈ d are the preset maximum change rates.

[0160] The adjusted parameters will be immediately applied to the next control cycle. This real-time adjustment mechanism enables the system to quickly adapt to changes in working conditions, such as load fluctuations, environmental temperature changes, etc., thus maintaining the best operating state.

[0161] The adaptive parameter adjustment step of the present invention greatly improves the flexibility and robustness of the energy management system. Compared with the control strategy with fixed parameters, this method can better cope with the complex and changeable working environment, extend the service life of the system, and improve the overall performance.

[0162] Finally, the method of the present invention also includes a model update step, which is the key to ensuring the long-term effectiveness of the control strategy. This step first collects the actual operation data of the system, including the measurement values of various physical quantities and the control input and output data. The data collection process uses high-frequency sampling (such as 100Hz) to capture the fast dynamic characteristics of the system.

[0163] The collected data is preprocessed to remove outliers and noise. The preprocessing methods include median filtering, wavelet transform, etc. The processed data is used for online parameter identification to update the system model. The present invention preferably uses the recursive least squares (RLS) method for parameter identification. Taking the fuel cell model as an example, its voltage output can be expressed as:

[0164] V fc = E0 - R·I - A·ln(I) - m·exp(n·I),

[0165] The core steps of the RLS algorithm include:

[0166] 1. Prediction:

[0167] 2. Calculate the prediction error:

[0168] 3. Calculate the gain vector: K(t) = P(t - 1)·φ(t)·[λ + φ T (t)·P(t - 1)·φ(t)] -1 ;

[0169] 4. Update the parameter estimate: θ(t) = θ(t - 1) + K(t)·e(t);

[0170] 5. Update the covariance matrix: P(t) = [λ -1 ·P(t - 1) - λ -1 ·K(t)·φ T (t)·P(t - 1)]; where, θ is the parameter vector, φ is the regression vector, and λ is the forgetting factor (usually taken as 0.95 - 0.99).

[0171] Through this method, the system model can be continuously updated to adapt to the slow changes in component performance (such as the aging of fuel cells). The updated model is then used for subsequent optimization of control strategies, forming a closed-loop adaptive process.

[0172] The model update step of the present invention ensures that the energy management system always makes decisions based on the latest and most accurate system model. This not only improves the control accuracy but also can timely detect abnormal changes in system performance, providing a basis for preventive maintenance. Combining the foregoing parameter optimization and adaptive adjustment steps, the method of the present invention constructs a comprehensive, intelligent, and adaptive energy management system, which can maintain the best performance under various working conditions, significantly improving the overall efficiency and reliability of the fuel cell lithium battery hybrid system.

[0173] It should be noted that the above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the principles of the present invention shall be included within the protection scope of the present invention.

Claims

1. Parameter optimization method for the energy management system of a fuel cell-lithium battery hybrid, characterized in that, Including: An acquisition step, including: Obtaining the working state data and load demand data of the hybrid power system; Obtaining the real-time performance parameters of the fuel cell and the lithium battery; A processing step, including: Determining the system operation mode based on the working state data and the load demand data; Executing the corresponding control strategy according to the system operation mode; Dynamically adjusting the parameters of the control strategy based on the real-time performance parameters; An output step, including: Generating an optimized power distribution instruction; Sending the power distribution instruction to the execution unit of the hybrid power system.

2. The method according to claim 1, wherein The control strategy includes at least one of a PID control strategy, a state machine decoupling control strategy, an ECMS equivalent hydrogen minimum consumption strategy, and an optimization control strategy based on DP-MPC.

3. The method according to claim 2, wherein The PID control strategy specifically includes: Obtaining an error signal between the system output power and the target power; Calculating the proportional term, integral term, and differential term based on the error signal; Generating a control signal according to the weighted sum of the proportional term, integral term, and differential term.

4. The method according to claim 2, wherein The state machine decoupling control strategy specifically includes: Determining the current state of the system based on the working state data; Selecting a preset power distribution strategy according to the state; Generating the power distribution ratio of the fuel cell and the lithium battery based on the preset power distribution strategy.

5. The method according to claim 2, wherein The ECMS equivalent hydrogen minimum consumption strategy specifically includes: Obtaining the instantaneous hydrogen consumption rate of the fuel cell and the equivalent hydrogen consumption rate of the lithium battery; Constructing an optimization objective function based on the instantaneous hydrogen consumption rate and the equivalent hydrogen consumption rate; Solving the optimization objective function to obtain an optimal power distribution scheme.

6. The method according to claim 2, wherein The optimization control strategy based on DP-MPC specifically includes: Predicting the load demand for a future period based on the current system state; Using a dynamic programming algorithm to solve the optimal control sequence within the prediction time domain; Executing the first control quantity of the optimal control sequence and repeating the optimization process in the next control cycle.

7. The method according to claim 1, characterized in that It also includes a parameter optimization step: Obtaining the system operation historical data; Constructing a system performance evaluation index based on the historical data; Using a multi-objective optimization algorithm to optimize the key parameters of the control strategy.

8. The method according to claim 7, wherein The system performance evaluation index includes at least one of hydrogen consumption, energy utilization rate, power stability, and system life.

9. The method according to claim 1, characterized in that It also includes an adaptive parameter adjustment step: Real-time monitoring of the system operation state and performance indicators; Dynamically adjusting the parameters of the control strategy based on the change trends of the operation state and performance indicators; Applying the adjusted parameters to the next control cycle.

10. The method according to claim 1, characterized in that, It also includes a model update step: Collecting the actual operation data of the system; Updating the system model using an online parameter identification method based on the actual operation data; Using the updated system model for subsequent control strategy optimization.

Images