Compound power energy management method and device based on dynamic weight self-adaptive adjustment

Abstract

This application discloses a method and device for hybrid power supply energy management based on dynamic weight adaptive adjustment, relating to the field of power supply energy management. The method includes: calculating the health status and lifetime loss rate of the hybrid power supply based on parameter data; calculating the dynamic weight of the hybrid power supply based on the health status; adjusting the dynamic weight of the hybrid power supply according to the total power demand to obtain the adjusted dynamic weight; constructing an optimization objective function based on the adjusted dynamic weight, lifetime loss rate, and total power demand; solving the optimization objective function using a quantum annealing algorithm under safety constraints to obtain a globally optimal power allocation result; and iteratively using a quantum approximation optimization algorithm and optimizer based on the globally optimal power allocation result to obtain a locally optimal power allocation result, thereby controlling the power of the hybrid power supply. This application can improve the energy utilization efficiency and lifetime of hybrid power supplies.

Description

Technical Field

[0001] This application relates to the field of power energy management technology, and in particular to a composite power energy management method and device based on dynamic weight adaptive adjustment. Background Technology

[0002] Hydrogen-electric hybrid power systems, as a key direction for next-generation power technologies, effectively address industry challenges such as short driving range and long charging times in pure electric vehicles, and slow dynamic response and high costs in pure fuel cell vehicles, by combining the high energy density of fuel cells with the high power density of lithium batteries. This hybrid architecture not only significantly improves the overall efficiency of the system but also demonstrates enormous potential in terms of vehicle power, economy, and reliability, and has become a mainstream technology in fields such as new energy vehicles, backup power, and distributed energy storage.

[0003] Despite the significant advantages of hybrid power systems, their energy management strategies still face serious challenges. In some cases, most technologies suffer from three core flaws: static design, singular objectives, and a lack of quantitative trade-off mechanisms. Specifically, control parameters remain fixed throughout the lifecycle of the hydrogen-electric hybrid power source, resulting in low energy utilization efficiency of both the hydrogen fuel cell and lithium battery. Furthermore, the lack of an effective online trade-off and coordination mechanism between the conflicting objectives of "protecting the hydrogen fuel cell" and "extending the lithium battery life" leads to a short lifespan for the hydrogen-electric hybrid power source. Summary of the Invention

[0004] The purpose of this application is to provide a method and device for energy management of composite power supplies based on dynamic weight adaptive adjustment, which can improve the energy utilization efficiency and lifespan of composite power supplies.

[0005] To achieve the above objectives, this application provides the following solution.

[0006] In a first aspect, this application provides a composite power supply energy management method based on dynamic weight adaptive adjustment, including the following:

[0007] Obtain parameter data and total power demand of the composite power source; the composite power source includes: fuel cell and lithium battery.

[0008] Based on the parameter data, the health status and lifespan loss rate of the composite power supply are calculated.

[0009] Based on the health status, the dynamic weight of the composite power source is calculated.

[0010] The dynamic weight of the composite power supply is adjusted according to the total power demand to obtain the adjusted dynamic weight of the composite power supply.

[0011] An optimization objective function is constructed based on the adjusted dynamic weights of the composite power supply, the lifetime loss rate, and the total power demand.

[0012] Under safety constraints, the quantum annealing algorithm is used to solve the optimization objective function to obtain the globally optimal power allocation result.

[0013] Based on the global optimal power allocation result, a quantum approximation optimization algorithm and optimizer are used for iteration to obtain the local optimal power allocation result.

[0014] The power of the composite power source is controlled based on the locally optimal power allocation result.

[0015] In a second aspect, this application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the composite power management method based on dynamic weight adaptive adjustment described in the first aspect.

[0016] According to the specific embodiments provided in this application, this application has the following technical effects.

[0017] This application calculates the health status and lifetime degradation rate of the hybrid power source using parameter data. Furthermore, it dynamically calculates the dynamic weight of the hybrid power source based on its health status, and then adjusts the dynamic weight again based on the total power demand, resulting in the adjusted dynamic weight of the hybrid voltage. An optimization objective function is then constructed using the adjusted dynamic weight, lifetime degradation rate, and total power demand. Finally, the global and local optimal power allocation results are obtained through quantum annealing and quantum approximation optimization algorithms, yielding the optimal power allocation for the fuel cell and lithium battery. This process not only achieves real-time dynamic adjustment of the weights of the fuel cell and lithium battery but also enables adaptive energy allocation, allowing the hybrid power source to dynamically balance lifetime protection and power balance, thereby improving its energy utilization efficiency and lifetime. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a flowchart illustrating a composite power management method based on dynamic weight adaptive adjustment, provided as an embodiment of this application.

[0020] Figure 2 This is a detailed flowchart illustrating a composite power management method based on dynamic weight adaptive adjustment, provided as an embodiment of this application.

[0021] Figure 3 for Figure 1 A detailed flowchart of steps S5 to S8.

[0022] Figure 4 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0023] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0024] The hybrid power energy management method based on dynamic weight adaptive adjustment provided in this application can be applied to hydrogen-electric hybrid power sources, such as vehicles that use fuel cells and lithium batteries to provide energy. By adaptively allocating energy between fuel cells and lithium batteries, the vehicle's hydrogen-electric hybrid power source can dynamically balance life protection and power balance, thereby improving the energy utilization efficiency and lifespan of the hybrid power source.

[0025] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0026] In one exemplary embodiment, such as Figure 1 and Figure 2 As shown, a composite power management method based on dynamic weight adaptive adjustment is provided. This method is executed by a computer device, specifically by a terminal or server alone, or by both a terminal and a server. In this embodiment, the method is described using a server as an example, and includes the following steps S1 to S8.

[0027] Step S1: Obtain the parameter data and total power demand of the hybrid power source; the hybrid power source includes: fuel cell and lithium battery.

[0028] As an feasible approach, the parameter data includes: fuel cell parameter data and lithium battery parameter data; fuel cell parameter data includes: output current, output voltage and stack temperature; lithium battery parameter data includes: charging current, discharging current, terminal voltage, battery temperature and state of charge.

[0029] Specifically, when applied to vehicles, fuel cells provide the primary power source, responsible for converting chemical energy into electrical energy. Lithium-ion batteries serve as auxiliary power sources, providing support during load fluctuations and energy recovery processes. A bidirectional DC / DC converter ensures the bidirectional flow of energy through power conversion between the fuel cell and lithium-ion batteries. Vehicle condition monitoring sensors, via communication networks (such as CAN bus or other compatible communication protocols), monitor and provide real-time feedback on the operating status of the fuel cell, lithium-ion batteries, and the vehicle. In addition to the parameters and total power demand of the combined power source, DC bus voltage is also available; fuel cell parameters include the cumulative number of start-stop cycles; and lithium-ion battery parameters include internal resistance.

[0030] Before performing step S2, the parameter data of the composite power source needs to be preprocessed, including: data validity verification (removing abnormal data that exceeds the reasonable physical range); noise filtering (eliminating high-frequency noise through a moving average filter with a window size of N); and time alignment (using timestamp alignment technology to ensure that all parameters have a unified time scale).

[0031] Step S2: Based on the parameter data, calculate the health status and lifetime loss rate of the composite power supply.

[0032] Specifically, after collecting parameter data for the hybrid power source, the health status of the fuel cell and lithium battery is assessed. By analyzing historical operating data and current status, the "aging degree" of the fuel cell and lithium battery can be quantitatively determined.

[0033] As an feasible approach, the health status of a hybrid power source includes the health status of the fuel cell and the health status of the lithium battery.

[0034] The formula for calculating the health status of a fuel cell is as follows.

[0035] .

[0036] in, for The health status of the fuel cell at all times for The health status of the fuel cell at all times The degradation coefficient of the fuel cell, for The output current of the fuel cell at all times This is the reference capacity for the fuel cell. The above formula represents... At any given moment, the health status of the fuel cell is updated based on the cumulative damage.

[0037] The formula for calculating the health status of a lithium battery is as follows.

[0038] .

[0039] .

[0040] .

[0041] .

[0042] in, for Monitor the health status of the lithium battery at all times. This represents the capacity decay coefficient of a lithium battery. for The cumulative capacity loss of lithium batteries over time This refers to the rated capacity of the lithium battery. for The cumulative capacity loss of lithium batteries over time for The incremental capacity loss of lithium batteries at all times This is a function of the lithium battery's degradation coefficient. for The state of charge of the lithium battery at all times. for The depth of discharge of the lithium battery at all times. for The temperature of the lithium battery at all times. for Flux at any given moment for The charging or discharging current of the lithium battery at all times. The sampling period.

[0043] Furthermore, this application can also calculate the efficiency of the fuel cell. and the internal resistance of lithium batteries As a basis for evaluating the performance status of fuel cells and lithium batteries.

[0044] As an feasible approach, the lifespan degradation rate of the hybrid power source includes the lifespan degradation rate of the fuel cell and the lifespan degradation rate of the lithium battery.

[0045] Specifically, the life loss rate of the composite power supply takes into account a variety of stress factors such as power fluctuations, operating current, and temperature deviation, reflecting the impact of the current operating state on the life of the fuel cell and lithium battery.

[0046] The formula for calculating the lifespan loss rate of a fuel cell is as follows.

[0047] .

[0048] in, for The lifespan and attrition rate of fuel cells at all times. , , and These are all weighting coefficients for fuel cells. for The power of the fuel cell at all times for The power of the fuel cell at all times This refers to the rated power of the fuel cell. for The output current of the fuel cell at all times The active area of ​​the fuel cell, for The temperature of the fuel cell stack is constantly monitored. The optimal operating temperature for fuel cells, for The output voltage of the fuel cell at all times This is the rated voltage of the fuel cell.

[0049] The formula for calculating the lifespan degradation rate of lithium batteries is as follows.

[0050] .

[0051] in, for The lifespan degradation rate of lithium batteries at all times. , , , and These are all weighting coefficients for lithium batteries. for The charging or discharging current of the lithium battery at all times. This refers to the rated capacity of the lithium battery. for The state of charge of the lithium battery at all times. This represents the optimal state of charge for lithium batteries. for The temperature of the lithium battery at all times. for The optimal operating temperature for lithium batteries at all times. for The depth of discharge of a lithium battery at any given time.

[0052] Step S3: Calculate the dynamic weight of the composite power source based on its health status.

[0053] Specifically, the core principle of calculating dynamic weights based on real-time health status is: the worse the power supply's health status, the higher the protection weight it receives. This means that power supplies with poor health status should be protected first to prevent them from aging too quickly.

[0054] As an feasible approach, the dynamic weighting of composite power sources includes dynamic weighting of fuel cells and dynamic weighting of lithium batteries.

[0055] The formula for calculating the dynamic weight of a fuel cell is as follows.

[0056] .

[0057] in, for Dynamic weighting of fuel cells at all times This is the adjustment factor for the fuel cell. for The health status of the fuel cell at all times.

[0058] The formula for calculating the dynamic weight of lithium batteries is as follows.

[0059] .

[0060] in, for Dynamic weighting of lithium batteries at all times. This is the adjustment coefficient for lithium batteries. for Monitor the health status of the lithium battery at all times.

[0061] Step S4: Adjust the dynamic weight of the composite power supply according to the total power demand to obtain the adjusted dynamic weight of the composite power supply.

[0062] As a feasible approach, if Then, increase the dynamic weight of the fuel cell and decrease the dynamic weight of the lithium battery to obtain the adjusted dynamic weights of the fuel cell and the lithium battery; the adjusted dynamic weight of the fuel cell is... The adjusted dynamic weight of lithium batteries is: ;like If the dynamic weights of the composite power source remain unchanged, then the dynamic weights of the composite power source remain unchanged; if Then, increase the dynamic weight of the lithium battery and decrease the dynamic weight of the fuel cell to obtain the adjusted dynamic weights of the fuel cell and lithium battery; the adjusted dynamic weight of the lithium battery is... The adjusted dynamic weights for fuel cells are: .

[0063] in, for Total power demand at any given time; High power threshold; The adjusted dynamic weights for fuel cells; This is the adjustment factor for the fuel cell. ; The adjusted dynamic weights for lithium batteries; This is the adjustment coefficient for lithium batteries. .

[0064] Step S5: Based on the adjusted dynamic weights of the composite power supply, lifetime loss rate, and total power demand, construct an optimization objective function.

[0065] As an feasible approach, the expression for optimizing the objective function is as follows.

[0066] .

[0067] .

[0068] .

[0069] .

[0070] in, for The objective function value at time t; The first weighting coefficient; for The lifespan loss rate of fuel cells at all times; This is the second weighting coefficient; for The lifespan degradation rate of lithium batteries at all times; This is the third weighting factor; for The power of the fuel cell at all times; for The power of the lithium battery at all times; for Total power demand at any given time; , and All are calibration coefficients; The adjusted dynamic weights for fuel cells; This refers to the adjusted dynamic weights of lithium batteries.

[0071] Specifically, , and It is a weighting coefficient calculated based on the adjusted dynamic weight of the composite power supply, used to reflect the protection priority of the composite power supply under different health conditions; This is the power error term, used to constrain the system power balance; , and Provided by simulation or bench calibration.

[0072] Step S6: Under security constraints, the quantum annealing algorithm is used to solve the optimization objective function to obtain the global optimal power allocation result.

[0073] Specifically, after constructing the objective function, the goal is to minimize the objective function value (i.e., minimize the total lifetime loss rate of the composite power source), and to solve the optimal power allocation strategy in real time under power balance and safety constraints. This application introduces a quantum computing-based hierarchical optimization mechanism, combining quantum annealing and quantum approximation optimization algorithms to achieve global and local cooperative optimal control, obtaining the optimal local optimal power allocation result.

[0074] As one possible implementation, step S6 specifically includes steps S61 to S68.

[0075] Step S61: Quantize the objective function to obtain the Hamiltonian model.

[0076] Specifically, the expression for the Hamiltonian model is as follows.

[0077] .

[0078] in, This represents the value of the Hamiltonian model.

[0079] Step S62: Determine the penalty term based on the safety constraints.

[0080] Specifically, the constraints are as follows.

[0081] Power balance constraints: .

[0082] Power constraints of fuel cells: .

[0083] Power constraints of lithium batteries: .

[0084] State of charge constraints: .

[0085] Temperature constraints: , .

[0086] Step S63: Construct the total Hamiltonian model based on the Hamiltonian model and the penalty term.

[0087] Specifically, the expression for the total Hamiltonian model is as follows.

[0088] .

[0089] in, The value of the total Hamiltonian model; This is a penalty item.

[0090] Step S64: Set the quantum annealing parameters and the initial Hamiltonian.

[0091] Step S65: Construct a quantum annealing algorithm model based on the quantum annealing parameters, initial Hamiltonian, and total Hamiltonian model.

[0092] Step S66: Perform annealing evolution on the quantum annealing algorithm model to obtain multiple global power allocation results.

[0093] Step S67: Substitute multiple global power allocation results into the optimization objective function to obtain multiple global objective function values.

[0094] Step S68: Determine the global power allocation result corresponding to the minimum global objective function value as the global optimal power allocation result.

[0095] Specifically, such as Figure 3 As shown, the quantum annealing algorithm is used to solve the optimization objective function, obtaining the globally optimal power allocation result within the global control cycle (such as second-level or task-level optimization). First, based on the optimization objective function superimposed on fuel cell lifetime loss and lithium battery lifetime loss, a penalty function is constructed in combination with constraints. The power allocation variables are discretized and mapped to qubit / spin variables, and finally, the total Hamiltonian including the constraints is constructed.

[0096] Subsequently, quantum annealing parameters such as total annealing time, number of annealing steps, and cross-field intensity evolution curve are set, an initial Hamiltonian is constructed, and its ground state is prepared, so that the algorithm is in a quantum superposition state covering all possible discrete values ​​of the power allocation variable, providing an initial state for subsequent quantum annealing in the entire power allocation search space.

[0097] Based on this, the algorithm follows Annealing evolution of quantum states, in which, For the first The Hamiltonian of an annealing step; These are the initial Hamiltonian coefficients; This is the initial Hamiltonian; This represents the total Hamiltonian coefficient. Decreasing Incremental approach, utilizing the quantum tunneling effect to gradually approximate from a high-energy quantum superposition state. The low-energy region is then determined. After annealing, multiple measurements and decodings are performed to obtain a set of candidate solutions that satisfy constraints such as power, state of charge, and temperature. The solution with the smallest global objective function value is selected as the globally optimal power allocation result for the global control cycle.

[0098] Step S7: Based on the global optimal power allocation result, the quantum approximation optimization algorithm and optimizer are used for iteration to obtain the local optimal power allocation result.

[0099] As one possible implementation, step S7 specifically includes steps S71 to S75.

[0100] Step S71: Based on the global optimal power allocation result, determine the initial circuit parameters of the quantum circuit function.

[0101] Specifically, the global optimal power allocation result is mapped to determine a qubit string of length q; the initial circuit parameters of the quantum circuit function are obtained by looking up the qubit string in a table (there are multiple initial circuit parameters, which are a set of numbers).

[0102] Step S72: Based on the quantum circuit function, the optimization objective function, and the optimizer, optimize the initial circuit parameters until the function value of the optimization objective function converges or the preset number of iterations is reached, and then obtain the optimal circuit parameters.

[0103] Specifically, the initial circuit parameters are substituted into the quantum circuit function to obtain multiple qubit strings; the multiple qubit strings are decoded to obtain multiple power allocation results; based on the multiple power allocation results, multiple objective function values ​​of the vehicle are calculated; based on the multiple objective function values, the average objective function value is calculated; the average objective function value is input into the optimizer to obtain updated circuit parameters; the updated circuit parameters are substituted into the quantum circuit function again to calculate the qubit strings iteratively until the average objective function value converges or the preset number of iterations is reached, thus obtaining the optimal circuit parameters.

[0104] Step S73: Input the optimal circuit parameters into the quantum circuit function, perform a preset number of calculations and decoding to obtain multiple local power allocation results.

[0105] Specifically, the optimal circuit parameters are input into the quantum circuit function, and multiple qubit strings are obtained through a preset number of calculations; the multiple qubit strings are then decoded to obtain multiple power allocation results.

[0106] Step S74: Based on multiple local power allocation results, use an optimization objective function to calculate multiple local objective function values.

[0107] Step S75: Select the local power allocation result corresponding to the minimum local objective function value and determine it as the local optimal power allocation result.

[0108] Specifically, a quantum approximation optimization algorithm and optimizer are used for iterative solution to obtain the local optimal power allocation result within a second-level control cycle (such as 1-second loop control). The quantum approximation optimization algorithm is then used to further refine the solution. A fast approximation optimization is then performed again. This algorithm constructs a parameterized quantum circuit function by alternately applying the target Hamiltonian and the hybrid Hamiltonian, and the expression for the quantum circuit function is as follows.

[0109] .

[0110] in, The value of the quantum circuit function; and These are circuit parameters, where the subscripts for the circuit parameters are 1, 2, 3... ;; ; ; For mixed operators; The initial quantum state prepared for the quantum circuit function at the beginning.

[0111] The average objective function value is input into the optimizer to obtain the updated circuit parameters, specifically including: inputting the circuit parameters... and Merge into a parameter vector: =[ , ..., , , ..., The average objective function value obtained by calculating and averaging is denoted as . , The optimization objective that needs to be minimized is... In each iteration, the average objective function value is input into the optimizer to obtain the updated circuit parameters, which specifically includes the following steps.

[0112] Step 1: Generate random perturbation direction: Generate and Random vectors of the same dimension Each of its components takes the value of or That is, for each circuit parameter (including each layer) and Randomly specify a fine-tuning direction: either "add a little" or "subtract a little". For example: =[ , , , ], =[+1,-1,+1,-1].

[0113] Step 2: Generate two sets of "trial parameters" with the disturbance amplitude set to... (This can be gradually reduced with iterations), based on the random direction in step 1, construct two sets of parameters: and .Right now, [ , , , ]and .

[0114] Step 3: Calculate the average objective function value for the two sets of parameters mentioned above. and By substituting the quantum circuit function into the quantum state and measuring the sample, two averaged objective function values ​​are obtained: .

[0115] Step 4: Determine the update direction and update the parameters based on "which side is smaller".

[0116] when Update along the positive direction. .

[0117] when Update along the negative direction. .

[0118] in, . This is the step size coefficient, which can be gradually decreased with each iteration to ensure convergence stability.

[0119] Step 5: If the change in the average objective function value has converged or reached the preset number of iterations, output the optimal circuit parameters; if the change in the average objective function value has not converged or has not reached the preset number of iterations, output the updated circuit parameters.

[0120] The globally optimal power allocation result obtained by the quantum annealing algorithm within the global cycle provides initial conditions for second-level quantum approximation optimization: At the boundary of the global cycle, the reference value of the fuel cell / lithium battery power allocation at the current moment is first extracted from the result, and the reference power allocation is mapped to an integer number according to the preset discretization and binary encoding rules, and then the number is written into a length of The binary bit string, which is mapped to The system assigns 0 / 1 values ​​to each qubit; during the offline phase, the system pre-calibrates the initial circuit parameters of a corresponding set of quantum circuit functions based on this bit encoding. Therefore, the current reference power allocation can be converted into the initial circuit parameters of the quantum circuit functions at the global periodic boundary.

[0121] Subsequently, within each second-level control cycle, the controller uses the circuit parameters obtained from the previous cycle's convergence as the starting point for the current cycle's iteration. Through quantum circuit functions: given this set of circuit parameters, the quantum circuit first encodes all candidate power allocation schemes. Each qubit is prepared as a quantum superposition state, with different qubit strings corresponding to different power allocation schemes and having different occurrence probabilities. Subsequently, the quantum circuit is executed multiple times, and measurements are taken on the qubits. Due to the randomness of quantum measurements, each measurement collapses to a specific qubit string according to the probability distribution. Therefore, multiple measurements under the same set of circuit parameters usually yield different qubit strings. The qubit strings obtained from each measurement are decoded one by one into candidate power allocation schemes, substituted into the optimization objective function to calculate the function value, and the average of these function values ​​is calculated to approximate the function value under the current circuit parameters. This function value is then fed back to the classical optimizer. The classical optimizer evaluates the merits of the current circuit parameters based on this, making small updates to the current circuit parameters within the same second-level control cycle, generating updated circuit parameters, and repeating the closed-loop iterative process of "running the quantum circuit—measuring—decoding—averaging—updating parameters" to gradually reduce the energy expectation. Once the iteration satisfies the convergence criterion or reaches the preset number of iterations, the optimal circuit parameters obtained in this cycle are fixed, and the quantum circuit is run and measured multiple times. From the candidate power allocation schemes that meet the constraints of power, state of charge, and temperature, the bit string with the smallest energy value is selected, decoded to obtain the local optimal power allocation result, and used as the actual execution instruction for this cycle. Thus, the tracking of the upper-level global optimal power allocation result and the real-time optimization control under dynamic operating conditions are realized in seconds.

[0122] Specifically, the quantum annealing algorithm and the quantum approximation optimization algorithm form a synergistic structure in this application: the quantum annealing algorithm is responsible for periodically searching for the globally optimal power allocation result, providing a reference strategy for the system over the entire lifetime; the quantum approximation optimization algorithm is responsible for solving the locally optimal power allocation result in real time within a second-level period, ensuring the system's immediate optimal response under dynamic operating conditions.

[0123] The two-layer algorithm operates collaboratively to achieve a balance between long-term lifetime optimization and short-term power balance: globally minimizing lifetime loss while locally ensuring rapid, stable, and efficient power output. In the hierarchical optimization collaborative mechanism of this application, the globally optimal power allocation result obtained by the quantum annealing algorithm, or its corresponding optimal allocation ratio, is used to characterize the target power load level of the fuel cell and lithium battery within the global control cycle, and is used to periodically calibrate control parameters such as dynamic weights and penalty factors, forming a power allocation reference strategy at the full lifetime scale. Within each second-level control cycle, the quantum approximation optimization algorithm, using the aforementioned locally optimal power allocation result and real-time load power as constraints and guidance, solves for the locally optimal power allocation result, which is then used as the actual control command for each power converter. Thus, the globally optimal power allocation result focuses on "long-term power allocation planning," while the locally optimal power allocation result prioritizes "instantaneous control execution." The two work collaboratively under a unified optimization model to achieve a balance between lifetime optimization and real-time power balance.

[0124] Step S8: Control the power of the composite power source based on the local optimal power allocation result.

[0125] Furthermore, the composite power supply energy management method based on dynamic weight adaptive adjustment provided in this application also includes step S9: comparing the parameter data of the composite power supply with a preset safety threshold, and when any parameter exceeds the limit, stopping the local optimal power allocation result to control the power of the composite power supply, and switching to the safety protection mode.

[0126] Specifically, after controlling the power of the composite power supply based on the local optimal power allocation result in step S8, the vehicle will compare any parameter data with a preset safety threshold. If the preset safety threshold is not exceeded, the power of the composite power supply will be controlled based on the local optimal power allocation result. If the preset safety threshold is exceeded, the vehicle will switch to the safety protection mode and control the power of the composite power supply based on the safe power of the safety protection mode.

[0127] Taking a vehicle as an example, in safety protection mode, based on the current parameters of the hybrid power source and key state quantities such as total power demand, the allowable safe power range of the fuel cell and lithium battery in this state is calculated by referring to a table or according to a preset relationship. Based on this, the original power allocation result is limited to obtain the safe power, which includes the safe power of the fuel cell. and lithium battery safety power and based on and The power of the combined power source is controlled. If the combined power of the fuel cell and lithium battery is insufficient to meet the load demand within the safe power range, a power or torque limit command is simultaneously sent to the vehicle controller to dredge the vehicle's output. Once the vehicle's monitoring module determines that the key state variables have returned to the safe range and remain there for a certain period, the controller automatically exits the safety protection mode and regenerates and sends control commands based on the locally optimal power allocation result according to the above method, thus achieving a smooth switch between safety protection and optimization control.

[0128] While completing safety assessments and issuing power commands, the system continuously inputs the historical optimal power allocation results obtained from quantum optimization, the deviation between the actual executed power and the reference power, constraint triggering and safety protection records, and corresponding key state variables into the vehicle's self-learning module to build an experience database. The self-learning module performs statistical analysis and fitting updates on the above historical data to correct key parameters such as adjustment coefficients in dynamic weight calculations, weighting coefficients in the objective function, and constraint penalty factors online, and appropriately calibrates some safety thresholds. This allows the quantum optimization model in subsequent control cycles to gradually conform to the actual degradation characteristics of the fuel cell stack and battery, achieving adaptive optimization of the control strategy.

[0129] The beneficial effects of the composite power supply energy management method based on dynamic weight adaptive adjustment proposed in this application are mainly reflected in three aspects.

[0130] 1. Treating fuel cells and lithium batteries as dual energy sources for collaborative optimization, a multi-objective real-time optimization problem is constructed with the goal of minimizing the system's weighted total lifetime loss rate. By introducing a quantum computing-based hierarchical optimization mechanism, a global optimal solution is achieved using quantum annealing, and a fast local real-time solution is achieved using a quantum approximation optimization algorithm. Combined with a dynamic weight adjustment strategy based on health status, the coupled problem of battery lifetime and power allocation is transformed into an energy allocation that can be adaptively and dynamically updated in each control cycle, thus realizing a paradigm shift from "static game optimization" to "quantum-driven adaptive intelligent optimization." This enables the system to dynamically balance lifetime protection and power balance, thereby improving the energy utilization efficiency and lifetime of the hybrid power source.

[0131] 2. The dynamic weight of the composite power supply was dynamically calculated based on its health status. The dynamic weight was then adjusted again based on the total power demand to obtain the adjusted dynamic weight of the composite voltage. This enabled adaptive energy allocation between the fuel cell and lithium battery under different health states, allowing the system to dynamically balance life protection and power balance, thereby improving energy utilization and operational reliability.

[0132] 3. Finally, the global optimal power allocation result and the local optimal power allocation result were solved by quantum annealing algorithm and quantum approximation optimization algorithm. The optimal power for energy allocation of fuel cell and lithium battery was obtained, realizing the coordinated power allocation of composite power supply at the whole life cycle and real-time control level, which significantly improved the optimization accuracy and real-time response capability of the system.

[0133] In one exemplary embodiment, a computer device is provided, which may be a server or a terminal, and its internal structure diagram may be as follows. Figure 4As shown, this computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operating system and computer programs in the non-volatile storage media to run. The database stores locally optimal power allocation results. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network connection. When the computer program is executed by the processor, it implements the aforementioned composite power management method based on dynamic weight adaptive adjustment.

[0134] Those skilled in the art will understand that Figure 4 The structures shown are merely block diagrams of some structures related to the present application and do not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than shown in the figures, or combine certain components, or have different component arrangements. In an exemplary embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0135] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0136] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0137] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0138] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0139] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for energy management of a compound power supply based on dynamic weight self-adaptive adjustment, characterized in that, The composite power supply energy management method based on dynamic weight adaptive adjustment includes: Obtain parameter data and total power demand of the hybrid power source; the hybrid power source includes: a fuel cell and a lithium battery; Based on the parameter data, the health status and lifespan loss rate of the composite power supply are calculated; Based on the health status, calculate the dynamic weight of the composite power source; The dynamic weight of the composite power supply is adjusted according to the total power demand to obtain the adjusted dynamic weight of the composite power supply. Based on the adjusted composite power supply dynamic weights, the lifetime loss rate, and the total power demand, an optimization objective function is constructed. Under safety constraints, the quantum annealing algorithm is used to solve the optimization objective function to obtain the globally optimal power allocation result; Based on the global optimal power allocation result, a quantum approximation optimization algorithm and optimizer are used for iteration to obtain the local optimal power allocation result; The power of the composite power source is controlled based on the locally optimal power allocation result.

2. The method of claim 1, wherein, The parameter data includes: fuel cell parameter data and lithium battery parameter data; the fuel cell parameter data includes: output current, output voltage and stack temperature; the lithium battery parameter data includes: charging current, discharging current, terminal voltage, battery temperature and state of charge.

3. The composite power supply energy management method based on dynamic weight adaptive adjustment according to claim 1, characterized in that, The health status of a hybrid power source includes the health status of the fuel cell and the health status of the lithium battery. The formula for calculating the health status of the fuel cell is as follows: ； in, for The health status of the fuel cell at all times for The health status of the fuel cell at all times The degradation coefficient of the fuel cell, for The output current of the fuel cell at all times This is the reference capacity for the fuel cell; The formula for calculating the health status of the lithium battery is as follows: ； ； ； ； in, for Monitor the health status of the lithium battery at all times. This represents the capacity decay coefficient of a lithium battery. for The cumulative capacity loss of lithium batteries over time This refers to the rated capacity of the lithium battery. for The cumulative capacity loss of lithium batteries over time for The incremental capacity loss of lithium batteries at all times This is a function of the lithium battery's degradation coefficient. for The state of charge of the lithium battery at all times. for The depth of discharge of the lithium battery at all times. for The temperature of the lithium battery at all times. for Flux at any given moment for The charging or discharging current of the lithium battery at all times. The sampling period.

4. The composite power supply energy management method based on dynamic weight adaptive adjustment according to claim 1, characterized in that, The lifespan attrition rate of a hybrid power source includes the lifespan attrition rate of the fuel cell and the lifespan attrition rate of the lithium battery. The formula for calculating the lifespan attrition rate of the fuel cell is as follows: ； in, for The lifespan and attrition rate of fuel cells at all times. , , and These are all weighting coefficients for fuel cells. for The power of the fuel cell at all times for The power of the fuel cell at all times This refers to the rated power of the fuel cell. for The output current of the fuel cell at all times The active area of ​​the fuel cell, for The temperature of the fuel cell stack is constantly monitored. The optimal operating temperature for fuel cells, for The output voltage of the fuel cell at all times This refers to the rated voltage of the fuel cell; The formula for calculating the lifespan degradation rate of the lithium battery is as follows: ； in, for The lifespan degradation rate of lithium batteries at all times. , , , and These are all weighting coefficients for lithium batteries. for The charging or discharging current of the lithium battery at all times. This refers to the rated capacity of the lithium battery. for The state of charge of the lithium battery at all times. This represents the optimal state of charge for lithium batteries. for The temperature of the lithium battery at all times. for The optimal operating temperature for lithium batteries at all times. for The depth of discharge of a lithium battery at any given time.

5. The composite power supply energy management method based on dynamic weight adaptive adjustment according to claim 1, characterized in that, The dynamic weight of the composite power source includes the dynamic weight of the fuel cell and the dynamic weight of the lithium battery. The formula for calculating the dynamic weight of the fuel cell is: ； in, for Dynamic weighting of fuel cells at all times This is the adjustment factor for the fuel cell. for Constantly monitor the health status of the fuel cell; The formula for calculating the dynamic weight of the lithium battery is as follows: ； in, for Dynamic weighting of lithium batteries at all times. This is the adjustment coefficient for lithium batteries. for Monitor the health status of the lithium battery at all times.

6. The composite power supply energy management method based on dynamic weight adaptive adjustment according to claim 5, characterized in that, The dynamic weight of the composite power supply is adjusted based on the total power demand to obtain the adjusted dynamic weight of the composite power supply, specifically including: like Then, the dynamic weight of the fuel cell is increased, and the dynamic weight of the lithium battery is decreased, resulting in the adjusted dynamic weights of the fuel cell and the lithium battery; the adjusted dynamic weight of the fuel cell is... The adjusted dynamic weight of the lithium battery is: ; like Then the dynamic weight of the composite power source remains unchanged; like Then, the dynamic weight of the lithium battery is increased, and the dynamic weight of the fuel cell is decreased, resulting in adjusted dynamic weights for the fuel cell and lithium battery; the adjusted dynamic weight of the lithium battery is... The adjusted dynamic weights of the fuel cell are: ; in, for Total power demand at any given time; High power threshold; The adjusted dynamic weights for fuel cells; This is the adjustment factor for the fuel cell. ; The adjusted dynamic weights for lithium batteries; This is the adjustment coefficient for lithium batteries. .

7. The composite power supply energy management method based on dynamic weight adaptive adjustment according to claim 1, characterized in that, The expression for the optimization objective function is: ； ； ； ； in, for The objective function value at time t; The first weighting coefficient; for The lifespan loss rate of fuel cells at all times; This is the second weighting coefficient; for The lifespan degradation rate of lithium batteries at all times; This is the third weighting factor; for The power of the fuel cell at all times; for The power of the lithium battery at all times; for Total power demand at any given time; , and All are calibration coefficients; The adjusted dynamic weights for fuel cells; This refers to the adjusted dynamic weights of lithium batteries.

8. The composite power supply energy management method based on dynamic weight adaptive adjustment according to claim 1, characterized in that, Under security constraints, the quantum annealing algorithm is used to solve the optimization objective function to obtain the globally optimal power allocation result, specifically including: The optimization objective function is quantized to obtain the Hamiltonian model; Based on the aforementioned security constraints, penalty terms are determined; Based on the Hamiltonian model and the penalty term, construct the total Hamiltonian model; Set the quantum annealing parameters and the initial Hamiltonian; Based on the quantum annealing parameters, the initial Hamiltonian, and the total Hamiltonian model, a quantum annealing algorithm model is constructed. The quantum annealing algorithm model is subjected to annealing evolution to obtain multiple global power allocation results; Substituting multiple global power allocation results into the optimization objective function yields multiple global objective function values; The global power allocation result corresponding to the minimum global objective function value is determined as the global optimal power allocation result.

9. The composite power supply energy management method based on dynamic weight adaptive adjustment according to claim 1, characterized in that, Based on the globally optimal power allocation result, a quantum approximation optimization algorithm and optimizer are used iteratively to obtain a locally optimal power allocation result, specifically including: Based on the global optimal power allocation result, the initial circuit parameters of the quantum circuit function are determined; Based on the quantum circuit function, the optimization objective function, and the optimizer, the initial circuit parameters are optimized until the function value of the optimization objective function converges or reaches a preset number of iterations, thus obtaining the optimal circuit parameters. The optimal circuit parameters are input into the quantum circuit function, and multiple local power allocation results are obtained through a preset number of calculations and decoding. Based on the multiple local power allocation results, the optimization objective function is used to calculate multiple local objective function values; The local power allocation result corresponding to the minimum local objective function value is determined as the local optimal power allocation result.

10. A computer device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program to implement the composite power management method based on dynamic weight adaptive adjustment as described in any one of claims 1-9.

Images