Wireless power supply system design method and system based on multi-target particle swarm optimization

Abstract

The invention provides a wireless power supply system design method based on multi-target particle swarm optimization, and the method comprises the steps: carrying out the characteristic analysis of each component of a bilateral LCC resonant topological structure wireless charging system, and determining a to-be-optimized component set; utilizing a Monte Carlo simulation method to simulate the influence degree of each component on the system performance when the parameter of each component to be optimized randomly fluctuates in a preset offset range; according to the quantification result of the influence degree, identifying one or more key resonance devices which have the greatest influence on the system performance; establishing a cost-performance multi-objective optimization function based on the key resonance device with the maximum influence; a multi-target particle swarm optimization algorithm is adopted to carry out collaborative optimization on parameter configuration of a key resonator device and a non-key resonator device, and a multi-target optimization optimal solution is given based on a Pareto frontier. According to the method, collaborative optimization of key and non-key resonator parameters is realized, and the efficiency and precision of parameter optimization are improved.

Description

Technical Field

[0001] This application relates to the field of wireless power transmission, and more specifically, to a design method for wireless power supply systems based on multi-target particle swarm optimization. Background Technology

[0002] Wireless power transfer (WPT) technology has been widely used in various applications due to its advantages such as safety, convenience, and space saving: consumer electronics, industrial automation, portable devices, and biomedical implants. The main function of the resonant network is to reduce reactive power and improve transmission efficiency. Different compensation topologies have been proposed and implemented to tune the resonant frequency in a wide range of applications. Based on the way compensation capacitors are added to the transmitting and receiving coils, there are four basic topologies: series-series (SS), series-parallel (SP), parallel-series (PS), and parallel-parallel (PP). To achieve high efficiency in WPT systems, some other complex topologies have also been proposed.

[0003] To reduce the size and cost of the additional inductor, a capacitor is connected in series with the primary winding to form an inductor-capacitor-capacitor (LCC) compensation network. Zero-current switching (ZCS) can be achieved by tuning the parameters of the LCC compensation network. The voltage and current stress of the LCC compensation topology is lower than that of other compensation topologies, and it is more efficient and stable.

[0004] Therefore, LCC compensation networks have received increasing attention in WPT systems in recent years. In real-world scenarios, various factors can affect the operation of wireless power transmission systems. Even slight misalignments between the transmitting and receiving coils can cause changes in the mutual inductance of the coupling coils. Environmental factors, such as temperature and aging, can also lead to variations in the component parameters of the compensation network.

[0005] In the prior art, patent CN117688886A discloses a method for optimizing the parameters of a wireless power transmission system compensation circuit. The method includes: constructing a wireless power transmission compensation circuit model based on a controlled source equivalent model of the compensation circuit; performing AC impedance analysis on the system to obtain expressions for the system output power and transmission efficiency; determining the target parameters and objective function for parameter optimization based on the expressions for system output power and transmission efficiency; determining constraints based on the requirements for system output power and transmission efficiency; and using an improved particle swarm optimization algorithm to optimize the parameters and obtain the optimal parameters for the wireless power transmission system compensation circuit. However, when multiple component parameters deviate simultaneously, it is difficult to determine which parameter has the greatest impact on system performance. This leads to an inability to balance component cost and system performance. Summary of the Invention

[0006] In view of one of the shortcomings of the prior art, the purpose of this application is to provide a design method for wireless power supply systems based on multi-target particle swarm optimization.

[0007] The first aspect of this application provides a design method for a dual-ended LCC-compensated wireless power supply system based on multi-target particle swarm optimization, comprising: The characteristics of each component in the bilateral LCC resonant topology wireless charging system are analyzed to determine the set of components to be optimized. The Monte Carlo simulation method is used to simulate the impact of each component on the system performance when the parameters of each component in the set of components to be optimized fluctuate randomly within a preset offset range. Based on the quantification of the degree of influence, one or more key resonant devices that have the greatest impact on system performance are identified from the set of components; Based on the key resonant device with the greatest impact, a cost-performance multi-objective optimization function is established. A multi-objective particle swarm optimization algorithm is used to collaboratively optimize the parameter configuration of the key resonant devices with high participation and the non-key resonant devices with low participation, generating multiple candidate solutions, and providing the optimal solution for multi-objective optimization based on the Pareto front.

[0008] Optionally, the set of components to be optimized includes: series resonant capacitors C1 and C2 on the primary and secondary sides; and parallel compensation capacitors C on the primary and secondary sides. P C S ; Primary and secondary series resonant inductance L P L S Primary and secondary coil inductances L1 and L2; The parameters of each of the aforementioned components satisfy the following: ; In the formula, This is the operating angular frequency.

[0009] Optionally, the step of using the Monte Carlo simulation method to simulate the impact of each component on system performance when the parameters of each component in the set of components to be optimized fluctuate randomly within a preset offset range includes: Obtain the components to be optimized, and build an accurate model of the WPT system using a two-sided LCC compensation network through simulation software. The simulation parameters are set according to the system parameters listed in Table I. The component to be optimized is treated as a random variable, and the parameter is randomly fluctuated according to a normal distribution within a set offset range of its nominal value. Generate a set of parameter combinations, run each set of parameters, and record and analyze the changes in the corresponding output power and transmission efficiency. Generate quantitative results of the degree of influence of each component on system performance.

[0010] Optionally, identifying one or more key resonant devices that have the greatest impact on system performance from the component set based on the quantification result of the degree of influence includes: Obtain the quantitative results of the degree of influence of each component on the system performance; Based on the quantification results, the components are sorted in order of maximum to minimum influence. Based on the sorting results, the two key resonant devices with the greatest impact were selected.

[0011] Optionally, the establishment of a cost-performance multi-objective optimization function based on the most influential key resonant device includes: Based on the critical resonant device with the greatest impact, different tolerance levels are selected; Construct a cost model and establish a cost correlation model determined by system performance changes and component parameters; Based on the aforementioned cost model and the established cost correlation model between system performance changes and component parameter conventions, an objective function related to overall cost and system stability is established. The expression for the objective function is: ; In the formula, To minimize system performance variations and component parameter costs; These are the calculated values ​​from the cost model; This represents the change in output power. Optionally, the cost model is an exponential cost model, and the expression for the exponential cost model is: ; In the formula, Base cost coefficient; T is the tolerance sensitivity coefficient; i Tolerance grade; The base of the exponential function; The expression for the cost correlation model determined by system performance changes and component parameters is as follows: ; In the formula, X i This is the nominal value of the component; P / X i This represents the sensitivity of the component to power.

[0012] Optionally, the multi-objective particle swarm optimization algorithm is used to collaboratively optimize the parameter configurations of the high-participation critical resonant devices and the low-participation non-critical resonant devices, generating multiple candidate solutions, and providing the optimal solution for multi-objective optimization based on the Pareto front, including: The quantitative results of the degree of influence are obtained, and each component is divided into critical resonant components and non-critical resonant components; Different constraint tolerances and different index codes are set according to the critical resonant devices and the non-critical resonant devices; Based on the multi-objective particle swarm optimization algorithm, the optimal parameter combination that simultaneously satisfies the requirements of performance stability and cost control is searched in parallel in the solution space, and multiple candidate solutions are generated. Based on the multiple candidate solutions, the Pareto front is used to filter the candidate solutions, and the optimal solution set of multi-objective optimization that takes into account both cost and performance is output.

[0013] Optionally, the constraint tolerance of the key resonant device is T. i Between 1% and 10%; The constraint tolerance of the non-critical resonant device is T. i ≥5%; The index codes for the critical resonant devices are real number codes, while the index codes for the non-critical resonant devices are E24 series codes.

[0014] A second aspect of this application provides a system for designing a wireless power supply system based on multi-objective particle swarm optimization, comprising: The component analysis module is used to perform characteristic analysis on each component of the bilateral LCC resonant topology wireless charging system and determine the set of components to be optimized. The quantization module is used to simulate the impact of each component on the system performance when the parameters of each component in the set of components to be optimized fluctuate randomly within a preset offset range using the Monte Carlo simulation method. An identification module is used to identify one or more key resonant devices that have the greatest impact on system performance from the set of components based on the quantification results of the degree of influence. A module is constructed to establish a cost-performance multi-objective optimization function based on the key resonant device with the greatest impact. The collaborative optimization module is used to perform collaborative optimization of the parameter configuration of the key resonant devices and non-key resonant devices using a multi-objective particle swarm optimization algorithm, and to provide the optimal solution for multi-objective optimization based on the Pareto front.

[0015] The wireless power supply system design method based on multi-objective particle swarm optimization provided in this application adopts the Monte Carlo method for designing the resonant element that has the most significant impact on system performance under multi-parameter deviation conditions, and uses the parameter design of the compensation network based on multi-objective particle swarm optimization. Based on the multi-objective particle swarm optimization algorithm, it improves the accuracy of the resonant device that has the greatest impact on system performance, thereby reducing the impact of device parameter deviation on system output power, while reducing the overall system cost. This solves the problem of balancing the sensitivity of resonant network parameters and manufacturing cost in the prior art.

[0016] Other technical effects resulting from the additional features will be further illustrated in the corresponding embodiments. Attached Figure Description

[0017] Other features, objects, and advantages of this application will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings: Figure 1 This is a flowchart illustrating a wireless power supply system design method based on multi-objective particle swarm optimization according to an exemplary embodiment; Figure 2 This is a simplified topology diagram of a wireless charging system based on a bilateral LCC compensation network according to an exemplary embodiment. Figure 3 The image shows a Monte Carlo simulation result according to an exemplary embodiment. Figure 4 This is a flowchart illustrating particle swarm optimization according to an exemplary embodiment; Figure 5 A diagram illustrating the Pareto front results according to an exemplary embodiment. Detailed Implementation

[0018] The present application will now be described in detail with reference to specific embodiments. These embodiments will help those skilled in the art to further understand the present application, but do not limit the present application in any way. It should be noted that those skilled in the art can make several modifications and improvements without departing from the concept of the present application, and these all fall within the protection scope of the present application. Parts not described in detail in the following embodiments can be implemented using existing technology.

[0019] In existing technologies, research on the impact of multi-parameter perturbations on system performance is insufficient. When multiple component parameters deviate simultaneously, it is difficult to determine which parameter has the greatest impact on system performance, leading to an inability to balance component cost and system performance. To address these issues, this application provides a wireless power supply system design method based on multi-objective particle swarm optimization to resolve these problems.

[0020] Reference Figure 1As shown in one embodiment of this application, a design method for a dual-ended LCC-compensated wireless power supply system based on multi-target particle swarm optimization includes: S1. Perform characteristic analysis on each component of the bilateral LCC resonant topology wireless charging system to determine the set of components to be optimized. S2. Using the Monte Carlo simulation method, simulate the impact of each component on the system performance when the parameters of each component in the set of components to be optimized fluctuate randomly within a preset offset range. S3. Based on the quantification results of the degree of influence, identify one or more key resonant devices that have the greatest impact on system performance from the component set; S4. Based on the key resonant components with the greatest impact, establish a cost-performance multi-objective optimization function; S5. A multi-objective particle swarm optimization algorithm is adopted to collaboratively optimize the parameter configuration of key resonant devices and other non-key resonant devices, and the optimal solution of multi-objective optimization is given based on the Pareto front.

[0021] It should be noted that multiple candidate solutions are generated through collaborative optimization. These solutions may have different values ​​in terms of cost and performance, and belong to the Pareto front. The optimal solution is then selected from these solutions.

[0022] Specifically, firstly, the characteristics of each component in the bilateral LCC resonant topology wireless charging system are analyzed to identify the set of components to be optimized. Then, using Monte Carlo simulation, the parameters of each component in the set to be optimized are simulated to exhibit random fluctuations within a preset offset range, quantifying the impact of each component on system performance. Next, based on the quantified impact results, one or more key resonant components with the greatest impact on system performance are selected from the component set. On this basis, a multi-objective optimization function balancing cost and performance is constructed around the identified key resonant components. Finally, a multi-objective particle swarm optimization algorithm is used to collaboratively optimize the parameter configurations of the key resonant components and other non-key resonant components. Ultimately, based on the Pareto front, the optimal solution for the multi-objective optimization is obtained, thus yielding a Pareto optimal solution set that characterizes the cost-performance trade-off.

[0023] The embodiments described above in this application accurately quantify the impact of component parameter fluctuations on system performance through Monte Carlo simulation, identify key resonant devices, and avoid blind optimization. By establishing a cost-performance multi-objective optimization function and combining it with a multi-objective particle swarm optimization algorithm, the collaborative optimization of key and non-key resonant device parameters is achieved. This solves the problem that single-objective optimization cannot simultaneously consider system performance and cost control, and improves the efficiency and accuracy of parameter optimization by leveraging the global search advantage of the particle swarm optimization algorithm. The final Pareto optimal solution set provides diverse parameter configuration options for practical engineering applications, allowing for flexible decision-making based on performance requirements and cost budgets in different scenarios, effectively improving the stability, reliability, and economy of the dual-end LCC compensated wireless power supply system.

[0024] In some specific embodiments of this application, the set of components to be optimized includes: series resonant capacitors C1 and C2 on the primary and secondary sides; and parallel compensation capacitors C on the primary and secondary sides. P C S ; Primary and secondary series resonant inductance L P L S The primary and secondary coil inductances are L1 and L2.

[0025] The WPT systems used in this application all employ bilateral LCC compensation networks. The simplified topology of the wireless charging system based on a bilateral LCC compensation network is as follows: Figure 2 As shown. Among them, u AB , u ab These are the inverter voltages for the primary and secondary sides, respectively. i 1. i 2 represents the current flowing through the primary coil and the secondary coil, respectively. L p , L s For series compensation inductance, C p , C s For parallel compensation capacitors, C 1. C 2 is a series compensation capacitor. L 1. L 2 represents the self-inductance of the primary coil and the secondary coil, respectively. M This represents the mutual inductance between the two coils. The primary-side resonant network consists of... L p With parallel branch ( C p +C 1) Constructed in series, the secondary resonant network consists of... L s With parallel branch ( C s+C 2) Connected in series.

[0026] The design requirements for the parameters of each component must meet the following: ; In the formula, This is the operating angular frequency.

[0027] In some specific embodiments of this application, when using the Monte Carlo simulation method to simulate the random fluctuation of the parameters of each component in the set of components to be optimized within a preset offset range, the degree of influence of each component on the system performance includes: The components to be optimized were obtained, and an accurate model of the WPT system using a two-sided LCC compensation network was built using simulation software. The simulation parameters were set according to the system parameters. The components to be optimized were treated as random variables, and the parameters were randomly fluctuated according to a normal distribution within the set offset range of their nominal values. A set of parameter combinations was generated, and each set of parameter combinations was run to record and analyze the changes in the corresponding output power and transmission efficiency. The quantitative results of the degree of influence of each component on the system performance were generated.

[0028] Specifically, the components to be optimized are first identified (primary and secondary series resonant capacitors C1 and C2, primary and secondary parallel compensation capacitors Cp and Cs, primary and secondary series resonant inductors Lp and Ls, and primary and secondary coil inductors L1 and L2). An accurate model of the wireless power transfer (WPT) system using a dual-sided LCC compensation network is built in the MATLAB / Simulink simulation environment, and the simulation parameters are configured based on the actual system parameters. Then, the components to be optimized are set as random variables, allowing their parameters to fluctuate randomly within a set offset range of their nominal values ​​according to a normal distribution. Next, multiple sets of parameter combinations are generated based on this random fluctuation rule. A simulation model is run once for each set of parameter combinations, and the changes in system output power and transmission efficiency corresponding to each simulation are recorded and analyzed simultaneously. Finally, through statistical analysis of these simulation data, a quantitative result is generated regarding the degree of influence of each component on system performance.

[0029] The simulation software used is the MATLAB / Simulink software environment.

[0030] The embodiments described above rely on a precise model of the dual-sided LCC-compensated WPT system built using MATLAB / Simulink, ensuring consistency between the simulation environment and the actual system. Simultaneously, a normal distribution is used to simulate the random fluctuations of component parameters, making the parameters closely match the natural deviation characteristics of component parameters in actual engineering, thus making the simulation scenario more realistic. By generating a large number of parameter combination sets and simulating them group by group, various fluctuations of component parameters within a preset offset range are comprehensively covered, avoiding the one-sidedness of single-parameter testing, overcoming the limitations of qualitative analysis, avoiding biases caused by subjective judgment, and improving the rationality of key component identification.

[0031] In some specific embodiments of this application, the set offset range is: ±10% of the nominal value.

[0032] The specific implementation method of this application is as follows: This Monte Carlo simulation aims to study the impact of parameter deviations of compensation components on the performance of a wireless power transfer (WPT) system. In the simulation, the parameters of eight compensation components (primary and secondary series resonant capacitors C1 and C2, primary and secondary parallel compensation capacitors Cp and Cs, primary and secondary series resonant inductors Lp and Ls, and primary and secondary coil inductors L1 and L2) change simultaneously within ±10% of their nominal values ​​according to a normal distribution, generating a total of 1000 parameter sets. The system automatically adjusts these component parameters within a specified deviation range and records the corresponding output power level under each parameter combination.

[0033] The simulation parameters of the system are shown in the table below:

[0034] In the simulation, a WPT system model with a two-sided LCC compensation network was built in the MATLAB / Simulink environment, and the simulation parameters were set according to the system parameters listed in Table I. Figure 5 Monte Carlo simulation results analyzing the impact of ±10% compensation component deviation on system output power are presented, where the vertical axis represents output power and the horizontal axis represents parameter tolerance. (Refer to...) Figure 3 As shown, the red and blue scatter plots correspond to values ​​higher and lower than the original component parameters, respectively. The yellow area represents the region where the output power exceeds the reference value, and the green area represents the region where the power level decreases. The scatter plots of the Cp and Cs parameters show a distinct right-skewed elliptical distribution, and the number of red clusters is significantly greater than that of blue points within the yellow area. In contrast, the distribution of the other six parameters is uniform and without directional bias, indicating that they are less sensitive to parameter changes.

[0035] In some specific embodiments of this application, the identification of one or more key resonant devices that have the greatest impact on system performance from the component set based on the quantification results of the degree of influence includes: Obtain the quantitative results of the influence of each component on the system performance; based on the quantitative results, sort the components in order of maximum to minimum influence; based on the sorting results, select the two key resonant devices with the greatest influence.

[0036] The embodiments described above in this application obtain the quantitative results of the influence of each component on the system output power and transmission efficiency generated by Monte Carlo simulation, and then sort the quantitative results from the degree of influence to the least. Finally, the two components with the most significant influence are selected from the sorted results as key resonant devices. This accurately locates the core components that play a decisive role in the performance of the bilateral LCC resonant topology wireless charging system, solving the problem in the prior art that it is difficult to identify key influencing factors when the parameters of multiple components deviate at the same time. This provides a clear and accurate optimization object for establishing a cost-performance multi-objective optimization function based on key resonant devices and carrying out targeted parameter optimization. It avoids the blindness of the optimization process, reduces the resource waste caused by the over-optimization of non-key components, and improves the system performance stability by focusing on key components, thereby achieving a balance between system performance and component cost.

[0037] In some specific embodiments of this application, the cost-performance multi-objective optimization function established based on the analysis results includes: Based on the quantitative results of the degree of impact, different tolerance levels are selected; Construct a cost model and establish a cost correlation model determined by system performance changes and component parameters; Based on the cost model and the cost correlation model between system performance changes and component parameter conventions, an objective function related to overall cost and system stability is established. The expression for the objective function is: ； In the formula, To minimize (system performance changes, component parameter costs); These are the calculated values ​​from the cost model; This refers to the change in output power.

[0038] Specifically, based on the quantitative results of the component impact obtained from the Monte Carlo simulation above, differentiated tolerance levels are selected for critical resonant components that have a significant impact on system performance and non-critical resonant components that have a smaller impact. Then, an exponential cost model C(T) is established for each. i (including base cost coefficient α) i Tolerance sensitivity coefficient β i With tolerance grade T i (correlation) and system performance changes (output power change ΔP) outA correlation model between the cost of components and their parameters was established. Finally, these two models were combined to construct a model that minimizes the total cost of components (f1=ΣC (T)). i ()) and minimizing system performance variation (f2=ΔP) out The overall objective function with the core objective being min (f1,f2)=(ΣC (T) i ),ΔP out By selecting components with differentiated tolerances to match the different impacts on system performance, cost waste or performance shortcomings caused by a single tolerance standard are avoided. The establishment of cost models and performance-cost correlation models realizes the quantitative binding of cost and performance. Combined with multi-objective functions, the core direction of optimization is clarified, solving the pain point of balancing component cost and system performance in existing technologies. It provides a quantifiable optimization basis for multi-objective particle swarm optimization algorithms, ensuring that the optimization process can simultaneously take into account the dual needs of system stability and cost control.

[0039] In some specific embodiments of this application, the cost model is an exponential cost model, and the expression for the exponential cost model is as follows: ; In the formula, Base cost coefficient; T is the tolerance sensitivity coefficient; i Tolerance grade; The base of the exponential function; The expression for the cost correlation model determined by system performance changes and component parameters is as follows: ; In the formula, X i This is the nominal value of the component; P / X i The sensitivity of a component to power; The total change in system output power.

[0040] Specifically, to achieve a balance between cost and performance, strategic component tolerance adjustment is crucial—using low-tolerance components for critical parameters while relaxing requirements for non-critical parameters. This paper proposes an algorithm to solve the tolerance allocation problem of WPT resonant networks, constructing a dual-objective optimization model regarding cost control and stability indices. Particle Swarm Optimization (PSO), as a population-based optimization algorithm, treats each particle as a potential solution. Particles navigate in the search space, updating their position and velocity by tracking individual best (pbest) and global best (gbest), aiming to minimize component costs while ensuring output power stability. The results are visualized using Pareto front analysis.

[0041] Objective functions are established for total cost and system stability respectively (denoted as f1 and f2).

[0042] . in, Minimize total cost; Minimal system performance change The exponential cost model is as follows: . In the formula, Base cost coefficient (related to components); This is the tolerance sensitivity coefficient; This refers to the tolerance level.

[0043] The expression for the cost correlation model determined by system performance changes and component parameters:

[0044] In the formula, Indicates the nominal value of the component; P / X i This indicates the sensitivity of a component to power.

[0045] In some specific embodiments of this application, the use of a multi-objective particle swarm optimization algorithm to collaboratively optimize the parameter configurations of critical resonant devices and other non-critical resonant devices, and to provide the optimal solution for multi-objective optimization based on the Pareto front, includes: Obtain quantitative results of the degree of impact, and classify each component as either critical or non-critical resonant components; Different constraint tolerances and different index codes are set according to the critical resonant devices and the non-critical resonant devices; Based on the multi-objective particle swarm optimization algorithm, the optimal parameter combination that simultaneously satisfies the requirements of performance stability and cost control is searched in parallel in the solution space, and multiple candidate solutions are generated. Based on multiple candidate solutions, the Pareto front is used to screen the candidate solutions and output the optimal solution set of multi-objective optimization that takes into account both cost and performance.

[0046] Specifically, refer to Figure 4 As shown, the process of using a multi-objective particle swarm optimization algorithm to collaboratively optimize parameters for high-involvement critical resonant devices and low-involvement non-critical resonant devices and generate Pareto optimal solution sets is based on the quantitative results of the component influence obtained from Monte Carlo simulation. First, based on these quantitative results, the components are divided into critical resonant devices (such as C) that have the greatest impact on system performance. P C SThe system first identifies two types of devices: critical resonators and non-critical resonators with less impact. Differential configurations are then assigned to these two types of devices. Critical resonators are subject to a 1% to 10% constraint tolerance and real-number coding, while non-critical resonators are subject to a ≥5% relaxed constraint tolerance and E24 series index coding to control costs and conform to industry standards. Subsequently, a particle population composed of critical and non-critical device parameters is initialized in the solution space. Each particle corresponds to a complete parameter configuration scheme. The total cost (f1) and performance fluctuation (f2=ΔP) of each particle are calculated by substituting them into an exponential cost model and a power variation correlation formula. out The fitness index is used to select the individual optimal (pbest) and global optimal (gbest) solutions. Then, the particle position and velocity are adjusted according to the continuous update rules of key parameters (including time-varying inertial weights and random perturbations) and the discrete index jump update rules of non-key parameters. After multiple rounds of iteration until the convergence condition is met, all parameter combinations that cannot improve one objective without worsening the other are extracted from the external archive set to form a Pareto optimal solution set that represents the cost-performance trade-off, providing diverse candidate solutions for engineering applications.

[0047] Among them, the multi-objective particle swarm optimization (MOPSO) algorithm in this application requires a customized strategy: firstly, because the key parameters have a significant impact on system performance, stricter tolerance constraints (T) need to be applied. i Constraints are set between 1% and 10% to ensure stability; non-critical parameters, due to their negligible impact, are subject to relaxed constraints (T). i (≥5%) to reduce costs and avoid over-allocation of resources. Through differentiated constraints, the algorithm prioritizes optimizing key parameters, reducing the total search space from 10¹ 0 Compress to 10 7 The convergence speed is improved by over 40%, and high-quality solution sets are obtained quickly. Secondly, the hybrid variable encoding mechanism improves algorithm efficiency while ensuring engineering feasibility: the tolerances of Cp and Cs are encoded with real numbers, allowing continuous adjustment within a 1% range; non-critical components are encoded with E24 series indexes, following the tolerance levels specified for standardized electronic components in the International Electrotechnical Commission standard IEC 60063, and are mapped to standard tolerance levels through indexes.

[0048] It should be noted that the core basis for classification is not the tolerance range, but the "quantification result of the degree of influence" of Monte Carlo simulation. By applying a ±10% normal distribution offset to the parameters of 8 types of components in the MATLAB / Simulink environment, the fluctuation range of output power and transmission efficiency is quantified. The components are sorted from largest to smallest fluctuation range, and the top two are selected as key resonant devices, while the rest are non-key devices. The tolerance range is a "subsequent constraint based on the classification result", not the basis for classification.

[0049] In some specific embodiments of this application, the constraint tolerance for the key resonant device is T. i Between 1% and 10%; the constraint tolerance for non-critical resonant components is T. i ≥5%; the index coding of critical resonant devices adopts real number coding, while the index coding of non-critical resonant devices adopts E24 series index coding.

[0050] In some specific embodiments of this application, the optimization rules for the parameter tolerances of key resonant devices are as follows: ; ; In the formula, Let d be the position of the t-th generation particle in d dimensions (dimension with continuously varying tolerances). This represents the update step size and direction in the d-dimensional dimension, which determines the parameter adjustment amount in subsequent iterations. Positive values ​​indicate an increase in tolerance, while negative values ​​indicate a decrease in tolerance. The time-varying inertial weights are used to balance global exploration; r1 and r2 are uniform random numbers used to introduce random perturbations. This represents the historical best tolerance value for a single particle in d dimensions; This represents the globally optimal tolerance value for all particles in the d-dimensional dimension. These are the variables for the tolerance parameters of key resonant devices.

[0051] Parameter tolerance update rules for non-critical resonant components: ; In the formula, Let be the index value of the t-th generation particle in dimension j; This represents the discrete velocity increment that controls the index jump magnitude; This is the E24 index value (historical candidate value) of the previous generation particle in dimension j.

[0052] Where E24(k) is the standardized tolerance range specified by the International Electrotechnical Commission (IEC 60063), and k is the range index; To find the gear index that minimizes the difference between the E24 standard tolerance and the iterative target value; To control the index jump range parameters (such as ±1, ±2), ensure that tolerance adjustments only switch between E24 standard ranges.

[0053] It should be noted that, for discrete processes, k in E24(k) is an index value, representing the "kth standard grade" in the E24 series. The core of the mapping is "assigning a unique sequence number k to each E24 standard value." The discrete process is essentially the expression for the tolerance update rule of non-critical parameters, through... Determine the corresponding value of the previous historical index, after The discrete increment is determined to obtain the continuous parameter. The corresponding standard value is retrieved through E24(k), and the deviation is calculated. The "E24 level with the smallest deviation" is selected, and the index k is determined. That is, the parameters of non-critical components are no longer represented by "continuous values" but by E24 series index encoding.

[0054] In this application, the improved particle swarm optimization algorithm refers to the inconsistent formulas used for optimizing high-participation elements and low-participation elements. Specifically, higher-participation elements are subject to stricter tolerance constraints, while low-participation elements are subject to relaxed constraints. This differentiated constraint approach improves the convergence speed and enables the rapid acquisition of high-quality solution sets.

[0055] Among them, key resonant devices are high-participation components, while non-key resonant devices are low-participation components.

[0056] The preferred features in the above embodiments can be used individually in any embodiment, or in any combination thereof, provided they do not conflict with each other. Furthermore, parts not described in detail in the embodiments can be implemented using existing technologies.

[0057] The foregoing has described some specific embodiments of this application. It should be understood that this application is not limited to the specific embodiments described above, and those skilled in the art can make various modifications or variations within the scope of the claims, which do not affect the substantive content of this application. The above-described preferred features can be used in any combination without conflict.

Claims

1. A design method for a wireless power supply system based on multi-objective particle swarm optimization, characterized in that, include: The characteristics of each component in the bilateral LCC resonant topology wireless charging system are analyzed to determine the set of components to be optimized. The Monte Carlo simulation method is used to simulate the impact of each component on the system performance when the parameters of each component in the set of components to be optimized fluctuate randomly within a preset offset range. Based on the quantification of the degree of influence, one or more key resonant devices that have the greatest impact on system performance are identified from the set of components; Based on the key resonant device with the greatest impact, a cost-performance multi-objective optimization function is established. A multi-objective particle swarm optimization algorithm is used to collaboratively optimize the parameter configurations of the key resonant devices and non-key resonant devices, and the optimal solution of the multi-objective optimization is given based on the Pareto front.

2. The wireless power supply system design method based on multi-objective particle swarm optimization according to claim 1, characterized in that, The set of components to be optimized includes: series resonant capacitors C1 and C2 on the primary and secondary sides; and parallel compensation capacitors C on the primary and secondary sides. P C S ; Primary and secondary series resonant inductance L P L S Primary and secondary coil inductances L1 and L2; The design requirements for the parameters of each component must meet the following: ； In the formula, This is the operating angular frequency.

3. The wireless power supply system design method based on multi-objective particle swarm optimization according to claim 1, characterized in that, The Monte Carlo simulation method is used to simulate the impact of each component on system performance when the parameters of each component in the set of components to be optimized fluctuate randomly within a preset offset range, including: Obtain the components to be optimized, and build an accurate model of the WPT system using a two-sided LCC compensation network through simulation software. The simulation parameters are set according to the system parameters. The component to be optimized is treated as a random variable, and the parameter is randomly fluctuated according to a normal distribution within a set offset range of its nominal value. Generate a set of parameter combinations, run each set of parameters, and record and analyze the changes in the corresponding output power and transmission efficiency. Generate quantitative results of the degree of influence of each component on system performance.

4. The wireless power supply system design method based on multi-objective particle swarm optimization according to claim 3, characterized in that, Based on the quantification results of the degree of influence, one or more key resonant devices that have the greatest impact on system performance are identified from the set of components, including: Obtain the quantitative results of the degree of influence of each component on the system performance; Based on the quantification results, the components are sorted in order of maximum to minimum influence. Based on the sorting results, the two key resonant devices with the greatest impact were selected.

5. The design method for a wireless power supply system based on multi-objective particle swarm optimization according to claim 1, characterized in that, Based on the key resonant device with the greatest impact, a cost-performance multi-objective optimization function is established, including: Based on the critical resonant components that have the greatest impact, different tolerance levels are selected; Construct a cost model and establish a cost correlation model determined by system performance changes and component parameters; Based on the cost model and the cost correlation model, an objective function related to overall cost and system stability is established. The expression for the objective function is: ； In the formula, To minimize system performance variations and component parameter costs; These are the calculated values ​​from the cost model; This represents the change in output power.

6. The design method for a wireless power supply system based on multi-objective particle swarm optimization according to claim 5, characterized in that, The cost model is an exponential cost model, and the expression for the exponential cost model is as follows: ； In the formula, Base cost coefficient; T is the tolerance sensitivity coefficient; i Tolerance grade; The base of the exponential function; The expression for the cost-related model is: ； In the formula, X i This is the nominal value of the component; P / X i This represents the sensitivity of the component to power.

7. The design method for a wireless power supply system based on multi-objective particle swarm optimization according to claim 1, characterized in that, The multi-objective particle swarm optimization algorithm is used to collaboratively optimize the parameter configurations of the critical and non-critical resonant devices, generating multiple candidate solutions, and providing the optimal solution based on the Pareto front, including: The quantitative results of the degree of influence are obtained, and each component is divided into critical resonant components and non-critical resonant components; Different constraint tolerances and different index codes are set according to the critical resonant devices and the non-critical resonant devices; Based on the multi-objective particle swarm optimization algorithm, the optimal parameter combination that simultaneously satisfies the requirements of performance stability and cost control is searched in parallel in the solution space, and multiple candidate solutions are generated. Based on the multiple candidate solutions, the Pareto front is used to filter the candidate solutions, and the optimal solution set of multi-objective optimization that takes into account both cost and performance is output.

8. The design method for a wireless power supply system based on multi-objective particle swarm optimization according to claim 7, characterized in that, The constraint tolerance of the key resonant device is T. i Between 1% and 10%; The constraint tolerance of the non-critical resonant device is T. i ≥5%; The index codes for the critical resonant devices are real number codes, while the index codes for the non-critical resonant devices are E24 series codes.

9. The design method for a wireless power supply system based on multi-objective particle swarm optimization according to claim 7, characterized in that, The parameter tolerance optimization rules for the key resonant device are as follows: ； ； In the formula, Let be the position of the t-th generation particle in d-dimensional space, with continuously varying tolerance dimensions. The update step size and direction in the d-dimensional dimension determine the parameter adjustment amount for subsequent iterations; a positive value indicates an increase in tolerance, and a negative value indicates a decrease in tolerance. The time-varying inertial weights are used to balance global exploration; r1 and r2 are uniform random numbers used to introduce random perturbations. This represents the historical best tolerance value for a single particle in d dimensions; This represents the globally optimal tolerance value for all particles in the d-dimensional dimension. The parameter tolerance optimization rules for non-critical resonant devices are as follows: ； In the formula, Let be the index value of the t-th generation particle in dimension j; This represents the discrete velocity increment that controls the index jump magnitude.

10. A system for designing a wireless power supply system based on multi-objective particle swarm optimization as described in any one of claims 1-9, characterized in that, include: The component analysis module is used to perform characteristic analysis on each component of the bilateral LCC resonant topology wireless charging system and determine the set of components to be optimized. The quantization module is used to simulate the impact of each component on the system performance when the parameters of each component in the set of components to be optimized fluctuate randomly within a preset offset range using the Monte Carlo simulation method. An identification module is used to identify one or more key resonant devices that have the greatest impact on system performance from the set of components based on the quantification results of the degree of influence. A module is constructed to establish a cost-performance multi-objective optimization function based on the key resonant device with the greatest impact. The collaborative optimization module is used to perform collaborative optimization of the parameter configuration of the key resonant devices and non-key resonant devices using a multi-objective particle swarm optimization algorithm, and to provide the optimal solution for multi-objective optimization based on the Pareto front.

Images