Unmanned aerial vehicle hydrogen battery efficiency optimization method and system
By dynamically optimizing the energy distribution and control parameters of the UAV hydrogen battery, the problem of inefficiency in the existing technology is solved, more efficient and reliable energy management is achieved, and operating risks and environmental impacts are reduced.
Patent Information
- Application Number
- CN202411095556.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-09
- Publication Date
- 2024-11-29
- Estimated Expiration
- 2044-08-09
AI Technical Summary
The existing drone hydrogen battery efficiency optimization technology is inefficient in handling real-time changes and dynamic responses under complex environmental conditions, and cannot effectively adjust energy output, resulting in waste of energy or insufficient energy supply, increasing system operation risks.
By collecting drone flight status and hydrogen battery data, a preliminary energy distribution model is built, the energy transmission path is dynamically optimized, the minimum cutting algorithm is introduced to identify key transmission nodes, and the differential evolution algorithm is used to optimize control parameters to ensure that the energy distribution matches the actual needs. Battery management.
It realizes more efficient and reliable energy management, reduces ineffective energy consumption, ensures the operating stability and reliability of the system at critical moments, extends battery life, and reduces environmental impact and operating costs.
Smart Images

Figure CN119045323A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of optimization control technology, and in particular to a method and system for optimizing the efficiency of hydrogen batteries for unmanned aerial vehicles. Background Art
[0002] Optimal control is a branch of automatic control engineering that focuses on developing and applying algorithms to improve the performance and efficiency of a system. This technique mainly uses various mathematical optimization methods, including linear and nonlinear programming, dynamic programming, and model-based predictive control. The goal of optimal control is to find the best control strategy so that the system operates at the best performance state, reduces energy consumption, and improves the reliability and stability of the system. It is widely used in multiple industries, including but not limited to energy systems, aerospace, automotive industry, and manufacturing.
[0003] Among them, the optimization method of drone hydrogen battery efficiency explores how to improve the efficiency of drone hydrogen batteries by optimizing control technology. This optimization method involves adjusting and improving the drone's energy management system to ensure that hydrogen batteries provide power while minimizing energy loss. This can not only extend the flight time of the drone, but also improve the overall energy utilization efficiency, so that the drone can operate more economically and efficiently when performing long-term surveillance, search and rescue or other tasks. In addition, improving the efficiency of hydrogen batteries can also help reduce operating costs and environmental impact, making drone technology an important step in sustainable development.
[0004] Although existing optimization control technologies are widely used in multiple industries, their application in optimizing the efficiency of hydrogen batteries for drones has shown certain limitations. Existing technologies mainly rely on traditional optimization methods, which are generally inefficient in dealing with real-time changes and dynamic responses under complex environmental conditions. When drones perform emergency tasks or face environmental changes, this inflexible optimization strategy may not be able to effectively adjust energy output, resulting in energy waste or insufficient energy supply. In addition, existing systems are often not powerful enough in fault prevention and rapid recovery, and cannot effectively identify key nodes or guarantee energy supply at critical moments, increasing the risk of system operation. These limitations not only increase operating costs, but also increase the environmental burden, affecting the potential application of drones in long-term and complex missions. The innovative solution effectively overcomes these shortcomings by introducing advanced dynamic adjustment technology and precise control algorithms, providing a more efficient and reliable solution for drone hydrogen battery management. Summary of the invention
[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a method and system for optimizing the efficiency of hydrogen batteries for unmanned aerial vehicles.
[0006] In order to achieve the above object, the present invention adopts the following technical solution: a method for optimizing the efficiency of a hydrogen battery of a drone, comprising the following steps:
[0007] S1: Collect the current flight status data of the drone, including speed and altitude, and the power data of the hydrogen battery, perform standardization, analyze and determine the energy requirements of each battery unit, and build a preliminary energy allocation model;
[0008] S2: dynamically optimizing the preliminary energy allocation model, reflecting changes in the energy transmission path by updating the graph model structure in real time, and adjusting the energy path according to sudden demands of the flight phase or additional tasks, matching the energy allocation with the actual demands of the UAV, and obtaining an adjusted energy allocation model;
[0009] S3: Based on the adjusted energy allocation model, a minimum cut algorithm is introduced to identify key transmission nodes and paths, prioritize energy supply to the key transmission nodes and paths, and isolate some battery units based on operating efficiency, dynamically optimize overall battery management, and obtain an optimized network energy allocation model;
[0010] S4: Based on the optimized network energy allocation model, the differential evolution algorithm is applied to optimize the parameters of the fuzzy controller, the shape and width of the membership function are adjusted, the adjusted control parameters are generated, and the discharge rate and power output of the battery are adjusted according to the real-time environment and task requirements.
[0011] As a further solution of the present invention, the steps of obtaining the preliminary energy allocation model are specifically as follows:
[0012] S111: Collect the speed, altitude and hydrogen battery power data of the drone using the formula:
[0013]
[0014] Generate standardized flight status and power data;
[0015] Where Z is the standardized flight status and power data, X represents the original data point, μ represents the mean value of the data, σ represents the standard deviation of the data, ∈ is the small amount, and γ is the adjustment factor;
[0016] S112: Based on the standardized flight status and power data, combined with the mission characteristics of the UAV, a multivariate linear regression model is used to predict the energy consumption of each battery unit, using the formula:
[0017] E=β 0 +β 1 V+β 2 H+β 3 V 2
[0018] Generates the energy requirements of each battery cell;
[0019] Where E represents energy demand, β θ , β 1 , β 2 , β 3 is the regression coefficient, V and H represent the standardized speed and height data, V 2 Increase the complexity of the model to match nonlinear changes;
[0020] S113: Utilizing the energy demand of each battery cell, applying a distribution formula to optimize energy distribution, using the formula:
[0021]
[0022] Generate a preliminary energy distribution model;
[0023] Among them, P i Represents the energy allocation percentage of the i-th battery cell, E i is the predicted energy demand of the ith battery cell, n is the total number of battery cells, and λ i is a factor adjusted according to the battery cell performance.
[0024] As a further solution of the present invention, the step of obtaining the adjusted energy allocation model is specifically as follows:
[0025] S211: Dynamically optimize the preliminary energy allocation model, update the graph model structure according to real-time flight data, and use the formula:
[0026]
[0027] Generate an updated graph model;
[0028] Where G represents the original graph model, Δt represents the time change since the last update, Δd represents the distance change since the last update, Δs represents the speed change, α, β, γ are the coefficients for adjusting the graph model update, G updated is the updated graph model;
[0029] S212: Use the updated graph model to recalculate the energy path using the formula:
[0030]
[0031] Generate optimized energy paths;
[0032] Among them, P new is the energy allocation model optimized based on the current flight status, G updated,jRepresents the energy transfer impact factor corresponding to the updated graph model;
[0033] S213: Match the optimized energy path with the actual demand of the UAV, adjust the energy path, and use the formula:
[0034] P final =min(P new , P demand )
[0035] Ensure that energy allocation does not exceed actual demand and obtain an adjusted energy allocation model;
[0036] Among them, P final is the adjusted energy allocation model.
[0037] As a further solution of the present invention, the step of obtaining the priority energy supply is specifically as follows:
[0038] S311: Based on the adjusted energy allocation model, a minimum cut algorithm is used to identify key transmission nodes and paths, using the formula:
[0039]
[0040] Generate a minimum cut set S;
[0041] Among them, C min represents the total capacity of the minimum cut, c ij represents the capacity of the edge in the graph, x ij represents the existence of the edge, λ i , j is the characteristic adjustment factor of the node;
[0042] S312: Extract key transmission nodes and paths according to the nodes and paths in the minimum cut set, using the formula:
[0043]
[0044] Generate key transmission node N key , and connect all key transmission nodes N key , establish the path set P key ;
[0045] Where V is the set of nodes in the graph, w ij represents the weight of the edge, θ is the threshold;
[0046] S313: For the key transmission node N key and the path set P key Prioritize energy supply and use the energy supply formula:
[0047]
[0048] Generate priority energy supply plan E priority ;
[0049] Among them, E priority Indicates the energy supplied first, E total represents the total energy, β key , β k represents the priority weight coefficient of the key node, and η is the sensitivity coefficient for adjusting energy distribution.
[0050] As a further solution of the present invention, the steps for obtaining the optimized network energy allocation model are specifically as follows:
[0051] S321: Calculate the operating efficiency of all battery cells using the formula:
[0052]
[0053] According to the operating efficiency of all battery cells, establish the efficiency ranking result R;
[0054] Among them, ∈ i is the operating efficiency of the battery cell, P i is the power output, E i is the energy consumption, ω i is the weight parameter, μ i is the adjustment parameter, σ i is a fine-tuning parameter, R is a list of battery cells sorted by operating efficiency;
[0055] S322: Isolate battery cells with efficiencies lower than a threshold value according to the efficiency ranking result, formula:
[0056] S low ={i∈R|∈ i <θ e ·κ-δ}
[0057] generating an isolated set of battery cells;
[0058] Among them, S low represents the isolated battery cell set, θ e is the basic threshold, κ is the adjustment threshold, and δ is the parameter of fine-tuning the threshold;
[0059] S323: Combine the key transmission node N key and the path set P key , and the battery cells that are not isolated, reconfigure the energy distribution using the formula:
[0060]
[0061] Generate an optimized network energy allocation model;
[0062] Among them, E net is the optimized network energy allocation model, α i is the distribution coefficient, ν i It is the sensitivity coefficient for adjusting the energy distribution of multiple battery cells.
[0063] As a further solution of the present invention, the step of obtaining the adjusted control parameters is specifically:
[0064] S411: Based on the optimized network energy allocation model, the differential evolution algorithm is applied to optimize the parameters of the fuzzy controller, using the formula:
[0065]
[0066] Generate a current control parameter set;
[0067] Among them, P opt is the current control parameter set, P best is the current optimal solution, P r1 , P r2 is a randomly selected solution, F is the difference weight, ΔE net is the change in the network energy allocation model, ∈ is the calculation stability parameter;
[0068] S412: According to the current control parameters, the shape and width of the fuzzy controller membership function are adjusted using the formula:
[0069] μ new =μ original ×(1+γ·log(1+|P opt |))
[0070] generating adjusted membership functions;
[0071] Among them, μ new is the adjusted membership degree, μ original is the original membership, γ is the adjustment coefficient, log(1+|P opt |) is a nonlinear function of parameter adjustment;
[0072] S413: Determine the fuzzy control parameters by combining the adjusted membership functions, using the formula:
[0073]
[0074] generating adjusted control parameters;
[0075] Among them, P control is the adjusted control parameter, μ new,i is the adjusted membership degree, wi is the weight coefficient.
[0076] A drone hydrogen battery efficiency optimization system, the drone hydrogen battery efficiency optimization system is used to execute the drone hydrogen battery efficiency optimization method, the system comprises:
[0077] The data collection module performs standardized processing and energy demand analysis based on the current flight status data of the drone, including speed and altitude, as well as the power data of the hydrogen battery, determines the energy demand of each battery unit, and builds a preliminary energy allocation model;
[0078] The dynamic optimization module is based on the preliminary energy allocation model, reflects the changes in the energy transmission path by updating the graph model structure in real time, generates a real-time optimization path, adjusts the energy path according to the sudden demand of the flight phase or additional tasks, matches the energy allocation with the actual demand of the UAV, and obtains the adjusted energy allocation model;
[0079] The key node identification module introduces a minimum cut algorithm based on the adjusted energy allocation model to identify key transmission nodes and paths, prioritizes energy supply to the key transmission nodes and paths, and isolates some battery units based on operating efficiency, dynamically optimizes overall battery management, and obtains an optimized network energy allocation model;
[0080] The control parameter optimization module optimizes the parameters of the fuzzy controller based on the optimized network energy distribution model, generates optimized control parameters, adjusts the shape and width of the membership function, and obtains adjusted control parameters.
[0081] Compared with the prior art, the advantages and positive effects of the present invention are:
[0082] In the present invention, by conducting a detailed analysis of the UAV flight status data and establishing an initial energy allocation model, the specific energy requirements of each battery unit can be met more accurately, thereby reducing ineffective energy consumption. The energy transmission path is updated in real time, and the dynamic adjustment of flight requirements is coordinated to ensure that energy allocation is always synchronized with actual requirements, thereby improving energy utilization efficiency. The minimum cut algorithm is introduced to optimize the energy supply of key nodes to ensure the operational stability and reliability of the system at critical moments. The control parameters optimized by the differential evolution algorithm allow fine adjustment of the battery discharge rate and power output to adapt to environmental changes and mission requirements, extend battery life and optimize energy management, reduce environmental impact and improve the economic benefits of the UAV. BRIEF DESCRIPTION OF THE DRAWINGS
[0083] Figure 1 It is a schematic diagram of the workflow of the present invention;
[0084] Figure 2A flowchart of the steps for obtaining the preliminary energy allocation model of the present invention;
[0085] Figure 3 A flow chart of the steps for obtaining the adjusted energy distribution model of the present invention;
[0086] Figure 4 A flow chart of the steps for obtaining the priority energy supply of the present invention;
[0087] Figure 5 A flowchart of the steps for obtaining the optimized network energy allocation model of the present invention;
[0088] Figure 6 This is a flow chart of the steps for obtaining the adjusted control parameters of the present invention. DETAILED DESCRIPTION
[0089] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.
[0090] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, "multiple" means two or more, unless otherwise clearly and specifically defined.
[0091] Embodiment 1
[0092] See also Figure 1 The present invention provides a technical solution: a method for optimizing the efficiency of a hydrogen battery for a drone, comprising the following steps:
[0093] S1: Collect the current flight status data of the drone, including speed and altitude, and the power data of the hydrogen battery, perform standardization, analyze and determine the energy requirements of each battery unit, and build a preliminary energy allocation model;
[0094] S2: Dynamically optimize the preliminary energy allocation model by updating the graph model structure in real time to reflect changes in the energy transmission path, and adjust the energy path according to the sudden demand of the flight phase or additional tasks to match the energy allocation with the actual needs of the UAV, and obtain the adjusted energy allocation model;
[0095] S3: Based on the adjusted energy allocation model, the minimum cut algorithm is introduced to identify key transmission nodes and paths, prioritize energy supply to key transmission nodes and paths, and isolate some battery units based on operating efficiency, dynamically optimize overall battery management, and obtain the optimized network energy allocation model;
[0096] S4: Based on the optimized network energy allocation model, the differential evolution algorithm is applied to optimize the parameters of the fuzzy controller, adjust the shape and width of the membership function, generate the adjusted control parameters, and adjust the battery discharge rate and power output according to the real-time environment and task requirements.
[0097] The preliminary energy distribution model specifically includes energy demand analysis, battery unit allocation ratio, and real-time data integration. The adjusted energy distribution model includes dynamic path adjustment, demand response matching, and update frequency. The optimized network energy distribution model includes a list of key nodes, optimized energy supply decisions, and isolation solutions. The adjusted control parameters include parameter optimization range, adjusted function characteristics, and control strategy accuracy.
[0098] See also Figure 2 , the specific steps for obtaining the preliminary energy allocation model are:
[0099] S111: Collect the speed, altitude and hydrogen battery power data of the drone using the formula:
[0100]
[0101] Generate standardized flight status and power data;
[0102] Where Z is the standardized flight status and power data, X represents the original data point, μ represents the mean value of the data, σ represents the standard deviation of the data, ∈ is the small amount, and γ is the adjustment factor;
[0103] S112: Based on the standardized flight status and power data, combined with the mission characteristics of the UAV, a multivariate linear regression model is used to predict the energy consumption of each battery cell, using the formula:
[0104] E=β 0 +β 1 V+β 2 H+β 3 V 2
[0105] Generates the energy requirements of each battery cell;
[0106] Where E represents energy demand, β 0 , β 1 , β 2 , β 3is the regression coefficient, V and H represent the standardized speed and height data, V 2 Increase the complexity of the model to match nonlinear changes;
[0107] S113: Utilizing the energy requirements of each battery cell, applying a distribution formula to optimize energy distribution, using the formula:
[0108]
[0109] Generate a preliminary energy distribution model;
[0110] Among them, P i Represents the energy allocation percentage of the i-th battery cell, E i is the predicted energy demand of the ith battery cell, n is the total number of battery cells, and λ i is a factor adjusted according to the battery cell performance.
[0111] Step 1: Data collection and standardization
[0112] Formula and parameters
[0113]
[0114] Parameter explanation:
[0115] X: original data point, assuming the speed data of the drone.
[0116] μ: The mean value of the data, calculated based on the collected data.
[0117] σ: Standard deviation of the data, calculated from the collected data.
[0118] ∈: A small amount, set to 1e-5 to avoid the denominator being zero.
[0119] γ: Adjustment factor used to enhance the sensitivity of standardization, generally set to 1.1.
[0120] Calculation process and examples
[0121] Assume that there are the following drone speed data points (unit: m / s): [10, 12, 15, 20, 25].
[0122] Calculate the average value μ: (10+12+15+20+25) / 5=16.4
[0123] Calculate the standard deviation σ:
[0124]
[0125] Use the formula to calculate the normalized result for a data point X=10:
[0126]
[0127] This normalized value Z represents the deviation of that data point from the average speed, taking into account the adjustment factor, which makes the results more sensitive to outliers.
[0128] Step 2: Energy Demand Forecast
[0129] Formula and parameters
[0130] E=β 0 +β 1 V+β 2 H+β 3 V 2
[0131] Parameter explanation:
[0132] E: predicted energy demand.
[0133] V, H: Normalized values of the drone’s speed and altitude, respectively.
[0134] β 0 , β 1 , β 2 , β 3 : Regression coefficient, obtained through statistical analysis of historical data.
[0135] Calculation process and examples
[0136] Assumptions:
[0137] The regression coefficient has been obtained through data analysis: β 0 =5,β 1 =0.5,β 2 =0.3,β 3 =0.05.
[0138] Normalized velocity V = -1.21 (obtained from step 1), height H = 0.5 (also obtained by normalization process).
[0139] Calculate energy requirements:
[0140] E=5+0.5×-1.21+0.3×0.5+0.05×(-1.21) 2
[0141] E=5-0.605+0.15+0.05×1.4641≈4.808
[0142] This value E represents the estimated energy demand of the drone in a given flight state, in units of energy (e.g., kilowatt-hours).
[0143] Step 3: Energy distribution model construction
[0144] Formula and parameters
[0145]
[0146] Parameter explanation:
[0147] P i : The energy allocation percentage of the ith battery cell.
[0148] E i : The predicted energy demand of the ith battery cell.
[0149] n: total number of battery cells.
[0150] λ i : Adjustment factor, set according to battery performance, assuming λ i =1.05.
[0151] Calculation process and examples
[0152] Assume there are three battery cells and the energy requirement of each cell is [4.808,5,6].
[0153] Calculate the denominator
[0154] ∑=4.808×1.05+5×1.05+6×1.05=15.858
[0155] For the first battery E 1 =4.808, distribution percentage:
[0156]
[0157] This percentage P 1 This means that the first cell should be allocated approximately 31.8% of the energy, adjusted for its performance.
[0158] See also Figure 3 , the specific steps for obtaining the adjusted energy allocation model are:
[0159] S211: Dynamically optimize the preliminary energy allocation model, update the graph model structure according to real-time flight data, and use the formula:
[0160]
[0161] Generate an updated graph model;
[0162] Where G represents the original graph model, Δt represents the time change since the last update, Δd represents the distance change since the last update, Δs represents the speed change, α, β, γ are the coefficients for adjusting the graph model update, Gupdated is the updated graph model;
[0163] S212: Use the updated graph model to recalculate the energy path using the formula:
[0164]
[0165] Generate optimized energy paths;
[0166] Among them, P new is the energy allocation model optimized based on the current flight status, G updated,j Represents the energy transfer impact factor corresponding to the updated graph model;
[0167] S213: Match the optimized energy path with the actual needs of the drone and adjust the energy path using the formula:
[0168] P final =min(P new , P demand )
[0169] Ensure that energy allocation does not exceed actual demand and obtain an adjusted energy allocation model;
[0170] Among them, P final is the adjusted energy allocation model.
[0171] Step 1: Update the graph model structure
[0172] formula:
[0173]
[0174] Parameter details:
[0175] G: Original graph model value. Assume the original value is 100.
[0176] Δt: Time change, for example, 5 minutes have passed since the last update, so let’s assume it’s 5.
[0177] Δd: distance change. For example, if the drone flies 2 kilometers, it is assumed to be 2.
[0178] Δs: speed change, for example, the speed increases from 10m / s to 15m / s, assuming the increase is 5m / s.
[0179] α, β, γ: adjustment coefficients, assuming α = 0.1, β = 0.05, γ = 0.02.
[0180] Calculation process:
[0181] Calculate the combined effects of time and distance:
[0182]
[0183] Update the graph model structure:
[0184] G updated =100×(1+0.545)=100×1.545=154.5
[0185] Step 2: Recalculate the energy transfer path
[0186] formula:
[0187]
[0188] Parameter details:
[0189] P i : Original energy distribution, assumed to be 30.
[0190] G updated : The value obtained from step 1 is 154.5.
[0191] n: total number of units, assuming 3, other units P j Assume both are 20.
[0192] Calculation process:
[0193] Calculate the denominator of the new energy path (assuming that all units have the same updated graph model):
[0194]
[0195] Calculate the molecular part of the new energy path:
[0196]
[0197] Update Energy Path:
[0198]
[0199] Step 3: Match energy allocation to actual demand
[0200] formula:
[0201] P final =min(P new , P demand )
[0202] Parameter details:
[0203] P new : The value calculated from step 2, which is 0.5.
[0204] P demand: Actual energy demand, assuming the drone demand is 0.4.
[0205] Calculation process:
[0206] Compare P new and P demand :
[0207] P final =min(0.5, 0.4)=0.4
[0208] Results
[0209] P final =0.4 means that the adjusted energy allocation model will be adjusted according to actual demand to ensure that the energy supply of the drone does not exceed the actual demand and optimize the energy utilization efficiency.
[0210] See also Figure 4 , the steps for obtaining priority energy supply are as follows:
[0211] S311: Based on the adjusted energy allocation model, the minimum cut algorithm is used to identify key transmission nodes and paths using the formula:
[0212]
[0213] Generate a minimum cut set S;
[0214] Among them, C min represents the total capacity of the minimum cut, c ij represents the capacity of the edge in the graph, x ij represents the existence of the edge, λ i , j is the characteristic adjustment factor of the node;
[0215] S312: Extract key transmission nodes and paths based on the nodes and paths in the minimum cut set, using the formula:
[0216]
[0217] Generate key transmission node N key , and connect all key transmission nodes N key , establish the path set P key ;
[0218] Where V is the set of nodes in the graph, wi j represents the weight of the edge, θ is the threshold;
[0219] S313: For key transmission node N key and the path set P key Prioritize energy supply and use the energy supply formula:
[0220]
[0221] Generate priority energy supply plan E priority ;
[0222] Among them, E priority Indicates the energy supplied first, E total represents the total energy, β key , β k represents the priority weight coefficient of the key node, and η is the sensitivity coefficient for adjusting energy distribution.
[0223] Step 1: Introduce the minimum cut algorithm
[0224] formula:
[0225]
[0226] Parameter interpretation and derivation:
[0227] c ij : The capacity of edge (i, j), which represents the maximum transmission capacity from node i to node j. Assume the value is 10.
[0228] x ij : The existence indicator function of edge (i, j), which is 1 if the edge exists, otherwise it is 0.
[0229] λ i , j : Characteristic coefficients of nodes i and j, these coefficients can represent the load or priority of the node, assuming λ i = 0.3 and λ j =0.5.
[0230] Calculation process:
[0231] Assume that there is an edge (i, j) in the network and this edge currently exists, that is, x ii =1.
[0232] Calculate the adjusted capacity of the edge:
[0233]
[0234] The value 5.56 represents the effective transmission capacity of edge (i, j) after considering node load and priority.
[0235] Step 2: Identify key transmission nodes and paths
[0236] formula:
[0237]
[0238] Parameter interpretation and derivation:
[0239] wi j : The weight of edge (i, j), which can be assigned based on traffic or importance, is assumed to be 8.
[0240] θ: weight threshold, used to filter important edges, assumed to be 5.
[0241] Calculation process:
[0242] For each edge (i, j), check whether its weight is greater than a threshold θ=5.
[0243] For the assumed w ij =8, since 8>5, edge (i, j) is considered to be the critical transmission path.
[0244] Therefore, nodes i and j are marked as critical transmission nodes.
[0245] Step 3: Prioritize energy supply
[0246] formula:
[0247]
[0248] Parameter interpretation and derivation:
[0249] E total : The total energy of the system, assumed to be 100 units.
[0250] β key , β k : Priority weight coefficient of key nodes, assuming β key =0.8, all key nodes β k The sum of is assumed to be 2.
[0251] η: Coefficient for adjusting the sensitivity of energy distribution, assumed to be 2.
[0252] Calculation process:
[0253] Calculate the total priority weight denominator:
[0254]
[0255] Calculate the priority energy supply ratio:
[0256]
[0257] Apply the regulation factor η to adjust the supplied energy:
[0258]
[0259] The value 6.32 represents the energy units allocated to the key nodes after adjustment under the given parameters.
[0260] See also Figure 5 ,The specific steps for obtaining the optimized network energy allocation model are:
[0261] S321: Calculate the operating efficiency of all battery cells using the formula:
[0262]
[0263] According to the operating efficiency of all battery cells, establish the efficiency ranking result R;
[0264] Among them, ∈ i is the operating efficiency of the battery cell, P i is the power output, E i is the energy consumption, ω i is the weight parameter, μ i is the adjustment parameter, σ i is a fine-tuning parameter, R is a list of battery cells sorted by operating efficiency;
[0265] S322: Based on the efficiency sorting result, isolate the battery cells whose efficiency is lower than the threshold, formula:
[0266] S low ={i∈R|∈ i <θ e ·κ-δ}
[0267] generating an isolated set of battery cells;
[0268] Among them, S low represents the isolated battery cell set, θ e is the basic threshold, κ is the adjustment threshold, and 6 is the parameter for fine-tuning the threshold;
[0269] S323: Combine key transmission node N key and the path set P key , and the battery cells that are not isolated, reconfigure the energy distribution using the formula:
[0270]
[0271] Generate an optimized network energy allocation model;
[0272] Among them, E net is the optimized network energy allocation model, α i is the distribution coefficient, μ i It is the sensitivity coefficient for adjusting the energy distribution of multiple battery cells.
[0273] Step 1: Calculate the overall battery cell operating efficiency
[0274] formula:
[0275]
[0276] P i : Power output of the battery unit (assumption: 100W)
[0277] ω i : Performance weighting coefficient (assumption: 1.2)
[0278] E i : Energy consumption of battery unit (assumption: 50W)
[0279] μ i : Energy consumption adjustment factor (assumption: 5W)
[0280] σ i : Fine-tuning parameter for stable calculation (assume: 0.5)
[0281] Calculation process
[0282] First calculate the numerator: (P i ×ω i ) 2 =(100×1.2) 2 =14400
[0283] Calculate the denominator: (E i +μ i ) 2 +σ i =(50+5) 2 +0.5=3025.5
[0284] Finally, calculate the efficiency:
[0285] Step 2: Isolate Inefficient Cells
[0286] formula:
[0287] S low ={i∈R|∈ i <θ e ·κ-δ}
[0288] θ e : Efficiency basic threshold (assumption: 3)
[0289] κ: Dynamic factor for adjusting the threshold (assumption: 1.1)
[0290] 6: Additional down-regulation factor (assumption: 0.2)
[0291] Calculation process
[0292] First calculate the threshold: θ eκ-δ=3×1.1-0.2=3.1
[0293] Compare the efficiency of each battery cell with the threshold: If ∈ i <3.1, then the battery cell i is added to S low In this example, ∈i≈4.76 and is therefore not isolated.
[0294] Step 3: Dynamically optimize overall battery management
[0295] formula:
[0296]
[0297] α i : Partition coefficient (assumption: 0.9)
[0298] V ν : Energy allocation sensitivity adjustment coefficient (assumption: 0.1)
[0299] P i : Power output of the battery cell (as before, 100W)
[0300] Calculation process
[0301] Calculate the allocated energy per battery cell:
[0302] The energy output of all unisolated batteries is accumulated to obtain E net .
[0303] See also Figure 6 , the specific steps for obtaining the adjusted control parameters are:
[0304] S411: Based on the optimized network energy allocation model, the differential evolution algorithm is used to optimize the parameters of the fuzzy controller using the formula:
[0305]
[0306] Generate a current control parameter set;
[0307] Among them, P opt is the current control parameter set, P best is the current optimal solution, P r1 , P r2 is a randomly selected solution, F is the difference weight, ΔE net is the change in the network energy allocation model, ∈ is the calculation stability parameter;
[0308] S412: According to the current control parameters, the shape and width of the fuzzy controller membership function are adjusted using the formula:
[0309] μ new =μ original ×(1+γ·log(1+|P opt |))
[0310] generating adjusted membership functions;
[0311] Among them, μ new is the adjusted membership degree, μ original is the original membership, γ is the adjustment coefficient, log(1+|P opt |) is a nonlinear function of parameter adjustment;
[0312] S413: Comprehensively adjust the membership function and determine the fuzzy control parameters using the formula:
[0313]
[0314] generating adjusted control parameters;
[0315] Among them, P control is the adjusted control parameter, μ new,i is the adjusted membership degree, w i is the weight coefficient.
[0316] Step 1: Apply differential evolution algorithm
[0317] formula:
[0318]
[0319] Parameter explanation and example derivation:
[0320] P best :The current optimal solution, set to P best =2.0
[0321] P r1 , P r2 : Two sets of control parameters are randomly selected, assuming that P r1 =3.5, P r2 =1.5
[0322] F: difference weight, controlling the degree of parameter variation, assuming F = 0.8
[0323] E net : The total value of the network energy distribution model, assumed to be 1000 units
[0324] ΔE net :E net The change in the amount, assuming it is 50 units
[0325] ∈: calculation stability parameter, used to prevent the denominator from being zero, set to 1
[0326] Derivation process:
[0327] Calculate P r1 -P r2 =3.5-1.5=2.0
[0328] Apply differential weights: F·(P r1 -P r2 )=0.8·2.0=1.6
[0329] calculate Final P opt =2.0+1.6+0.04995≈3.65
[0330] Explanation of calculation results:
[0331] P opt It is approximately 3.65, which indicates the parameter value of the candidate solution newly generated in this iteration.
[0332] Step 2: Adjust the shape and width of the membership function
[0333] formula:
[0334] μ new =μ original ×(1+γ·log(1+|P opt |))
[0335] Parameter explanation and example derivation:
[0336] μ original : Original membership, assumed to be 0.5
[0337] P opt : 3.65 calculated in the previous step
[0338] γ: adjustment coefficient, assumed to be 0.3
[0339] Derivation process:
[0340] Calculate log(1+|P opt |)=log(1+3.65)≈1.5
[0341] Calculate γ·log(1+|P opt |)=0.3·1.5=0.45
[0342] Final μ new =0.5×(1+0.45)=0.5×1.45=0.725
[0343] Explanation of calculation results:
[0344] μ new It is 0.725, which means that after adjustment, the membership degree increases and the membership function becomes wider or smoother.
[0345] Step 3: Generate adjusted control parameters
[0346] formula:
[0347]
[0348] Parameter explanation and example derivation:
[0349] μ new,i : The adjusted membership degree, assuming there are two control parameters, both are 0.725
[0350] w i : Parameter weights, assumed to be 0.6 and 0.4
[0351] Derivation process:
[0352] For each i:
[0353] First parameter:
[0354] Second parameter:
[0355] Final P control =0.66+0.54=1.2
[0356] Explanation of calculation results:
[0357] P control The value of is 1.2, which means the control parameter value after comprehensive consideration of the adjusted membership and weight, which is used in fuzzy control systems to improve the response effect and accuracy of the system.
[0358] A drone hydrogen battery efficiency optimization system, which is used to execute the drone hydrogen battery efficiency optimization method, and the system includes:
[0359] The data collection module performs standardized processing and energy demand analysis based on the current flight status data of the drone, including speed and altitude, as well as the power data of the hydrogen battery, determines the energy demand of each battery unit, and builds a preliminary energy allocation model;
[0360] The dynamic optimization module is based on the preliminary energy allocation model. It updates the graph model structure in real time to reflect the changes in the energy transmission path, generates a real-time optimization path, adjusts the energy path according to the sudden demand of the flight phase or additional tasks, matches the energy allocation with the actual needs of the UAV, and obtains the adjusted energy allocation model.
[0361] The key node identification module introduces the minimum cut algorithm based on the adjusted energy allocation model to identify key transmission nodes and paths, prioritize energy supply to key transmission nodes and paths, and isolate some battery units based on operating efficiency, dynamically optimize overall battery management, and obtain an optimized network energy allocation model;
[0362] The control parameter optimization module optimizes the parameters of the fuzzy controller based on the optimized network energy distribution model, generates optimized control parameters, adjusts the shape and width of the membership function, and obtains the adjusted control parameters.
[0363] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.
Claims
1. A method for optimizing the efficiency of hydrogen batteries for unmanned aerial vehicles, characterized in that: The following steps are involved: Collect the current flight status data of the drone, including speed and altitude, and the power data of the hydrogen battery, perform standardization, analyze and determine the energy requirements of each battery unit, and build a preliminary energy distribution model; Dynamically optimize the preliminary energy allocation model, reflect changes in the energy transmission path by updating the graph model structure in real time, and adjust the energy path according to sudden demands of the flight phase or additional tasks, so as to match the energy allocation with the actual needs of the UAV, and obtain an adjusted energy allocation model; Based on the adjusted energy allocation model, a minimum cut algorithm is introduced to identify key transmission nodes and paths, prioritize energy supply to the key transmission nodes and paths, and isolate some battery cells based on operating efficiency, dynamically optimize overall battery management, and obtain an optimized network energy allocation model; Based on the optimized network energy allocation model, the differential evolution algorithm is applied to optimize the parameters of the fuzzy controller, adjust the shape and width of the membership function, generate adjusted control parameters, and adjust the discharge rate and power output of the battery according to the real-time environment and task requirements.
2. The method for optimizing the efficiency of hydrogen batteries for unmanned aerial vehicles according to claim 1, characterized in that: The steps for obtaining the preliminary energy allocation model are specifically as follows: Collect the speed, altitude and hydrogen battery power data of the drone, using the formula: Generate standardized flight status and power data; Among them, Z is the standardized flight status and power data, X represents the original data point, μ represents the mean value of the data, σ represents the standard deviation of the data, ∈ is the small amount, and Y is the adjustment factor; Based on the standardized flight status and power data, combined with the mission characteristics of the UAV, a multivariate linear regression model is used to predict the energy consumption of each battery cell, using the formula: E=β0+β1V+β2H+β3V 2 Generates the energy requirements of each battery cell; Among them, E represents energy demand, β0, β1, β2, β3 are regression coefficients, V and H represent the standardized speed and height data, and V 2 Increase the complexity of the model to match nonlinear changes; Using the energy requirements of each battery cell, an allocation formula is applied to optimize energy allocation, using the formula: Generate a preliminary energy distribution model; Among them, P i Represents the energy allocation percentage of the i-th battery cell, E i is the predicted energy demand of the ith battery cell, n is the total number of battery cells, and λ i is a factor adjusted according to the battery cell performance.
3. The method for optimizing the efficiency of hydrogen batteries for unmanned aerial vehicles according to claim 2, characterized in that: The steps for obtaining the adjusted energy allocation model are specifically as follows: The preliminary energy allocation model is dynamically optimized, and the graph model structure is updated according to real-time flight data, using the formula: Generate an updated graph model; Where G represents the original graph model, Δt represents the time change since the last update, Δd represents the distance change since the last update, Δs represents the speed change, α, β, γ are the coefficients for adjusting the graph model update, G updated is the updated graph model; Using the updated graph model, the energy path is recalculated using the formula: Generate optimized energy paths; Among them, P new is the energy allocation model optimized based on the current flight status, G updated,j Represents the energy transfer impact factor corresponding to the updated graph model; Match the optimized energy path with the actual needs of the drone and adjust the energy path using the formula: P final =min(P new ,P demand ) Ensure that energy allocation does not exceed actual demand and obtain an adjusted energy allocation model; Among them, P final is the adjusted energy allocation model.
4. The method for optimizing the efficiency of hydrogen batteries for unmanned aerial vehicles according to claim 3, characterized in that: The steps for obtaining the priority energy supply are specifically as follows: Based on the adjusted energy allocation model, the minimum cut algorithm is used to identify key transmission nodes and paths, using the formula: Generate a minimum cut set S; Among them, C min represents the total capacity of the minimum cut, c ij represents the capacity of the edge in the graph, x ij represents the existence of the edge, λ i , j is the characteristic adjustment factor of the node; According to the nodes and paths in the minimum cut set, the key transmission nodes and paths are extracted using the formula: Generate key transmission node N key , and connect all key transmission nodes N key , establish the path set P key ; Where V is the set of nodes in the graph, w ij represents the weight of the edge, θ is the threshold; For the key transmission node N key and the path set P key Prioritize energy supply and use the energy supply formula: Generate priority energy supply plan E priority ; Among them, E priority Indicates the energy supplied first, E total represents the total energy, β key , β k represents the priority weight coefficient of the key node, and η is the sensitivity coefficient for adjusting energy distribution.
5. The method for optimizing the efficiency of hydrogen batteries for unmanned aerial vehicles according to claim 4, characterized in that: The steps for obtaining the optimized network energy allocation model are specifically as follows: Calculate the operating efficiency of all battery cells using the formula: According to the operating efficiency of all battery cells, establish the efficiency ranking result R; Among them, ∈ i is the operating efficiency of the battery cell, P i is the power output, E i is the energy consumption, ω i is the weight parameter, μ i is the adjustment parameter, σ i is the fine-tuning parameter, and R is the list of battery cells sorted according to operating efficiency; According to the efficiency ranking results, the battery cells with efficiency lower than the threshold are isolated, and the formula is: S low ={i∈R|∈ i <θ e ·k-d} generating an isolated set of battery cells; Among them, S low represents the isolated battery cell set, θ e is the basic threshold, K is the adjustment threshold, and δ is the parameter for fine-tuning the threshold; Combined with the key transmission node N key and the path set P key , and the battery cells that are not isolated, reconfigure the energy distribution using the formula: Generate an optimized network energy allocation model; Among them, E net is the optimized network energy allocation model, α i is the distribution coefficient, ν i It is the sensitivity coefficient for adjusting the energy distribution of multiple battery cells.
6. The method for optimizing the efficiency of hydrogen batteries for unmanned aerial vehicles according to claim 5, characterized in that: The steps for obtaining the adjusted control parameters are specifically as follows: Based on the optimized network energy allocation model, the differential evolution algorithm is applied to optimize the parameters of the fuzzy controller using the formula: Generate a current control parameter set; Among them, P opt is the current control parameter set, P best is the current optimal solution, P r1 , P r2 is a randomly selected solution, F is the difference weight, ΔE net is the change in the network energy allocation model, ∈ is the calculation stability parameter; According to the current control parameters, the shape and width of the fuzzy controller membership function are adjusted using the formula: m new =μ original ×(1+γ·log(1+|P opt |)) generating adjusted membership functions; Among them, μ new is the adjusted membership degree, μ original is the original membership, γ is the adjustment coefficient, log(1+|P opt |) is a nonlinear function of parameter adjustment; Based on the adjusted membership function, the fuzzy control parameters are determined using the formula: generating adjusted control parameters; Among them, P control is the adjusted control parameter, μ new,i is the adjusted membership degree, w i is the weight coefficient.
7. A drone hydrogen battery efficiency optimization system, characterized in that: According to the method for optimizing the efficiency of a hydrogen battery for a drone according to any one of claims 1 to 6, the system comprises: The data collection module performs standardized processing and energy demand analysis based on the current flight status data of the drone, including speed and altitude, as well as the power data of the hydrogen battery, determines the energy demand of each battery unit, and builds a preliminary energy allocation model; The dynamic optimization module is based on the preliminary energy allocation model, reflects the changes in the energy transmission path by updating the graph model structure in real time, generates a real-time optimization path, adjusts the energy path according to the sudden demand of the flight phase or additional tasks, matches the energy allocation with the actual demand of the UAV, and obtains the adjusted energy allocation model; The key node identification module introduces a minimum cut algorithm based on the adjusted energy allocation model to identify key transmission nodes and paths, prioritizes energy supply to the key transmission nodes and paths, and isolates some battery units based on operating efficiency, dynamically optimizes overall battery management, and obtains an optimized network energy allocation model; The control parameter optimization module optimizes the parameters of the fuzzy controller based on the optimized network energy distribution model, generates optimized control parameters, adjusts the shape and width of the membership function, and obtains adjusted control parameters.
Citation Information
Patent Citations
Unmanned aerial vehicle hybrid propulsion system energy management and control method based on flight data
CN110348595A
Evaluation method for hybrid power supply architecture of new energy unmanned aerial vehicle
CN112060983A
Hydrogen fuel unmanned aerial vehicle energy management method based on deep reinforcement learning
CN113200148A
Hydrogen energy unmanned aerial vehicle control system and method based on model prediction algorithm
CN116909306A
Unmanned aerial vehicle laser charging efficiency optimization method and system
CN118107826A
Cited By
Electrical equipment and control system thereof
CN119828476A
Unmanned aerial vehicle power management method, device and system based on thermal stealth material
CN119902457A
Thermal-invisibility-material-based unmanned aerial vehicle power management method, device and system
CN119902457B
