An improved Grey Wolf algorithm parameter identification method based on wireless power transmission system

Abstract

The present invention proposes a parameter identification method based on an improved Gray Wolf algorithm for a wireless power transmission system. The wireless power transmission system of the present invention includes a DC power supply, an inverter circuit module, a transmitting end resonant compensation capacitor, a transmitting coil inductance, a transmitting coil equivalent resistance, a receiving end resonant compensation capacitor, a receiving coil inductance, a receiving coil equivalent resistance, and a load resistance. The method of the present invention combines a wireless power transmission efficiency model to construct a parameter identification target model, and uses an improved Gray Wolf algorithm to transform the load and mutual inductance identification problem of the wireless power transmission system into a problem of solving the minimum error between the actual detection value and the calculated value of the transmitting end resonant current, and optimally identifies the load resistance and mutual inductance value with the minimum error value. The present invention effectively reduces circuit complexity and system volume, and can prevent the Gray Wolf algorithm from falling into a local optimal solution, so as to better design the wireless power transmission system as a whole.

Description

Technical Field

[0001] The present invention belongs to the technical field of wireless power transmission, and in particular relates to an improved grey wolf algorithm parameter identification method based on a wireless power transmission system. Background Art

[0002] Wireless Power Transfer (WPT) is a novel power transmission technology. With the continuous advancement of the world, people are using power equipment in a variety of complex situations. While traditional wired power transmission can meet most power supply needs, its shortcomings are becoming increasingly apparent in certain applications, such as safety issues caused by aging transmission wires. Therefore, wireless power transmission, due to its safety and flexibility, has become a new research hotspot.

[0003] Wireless power transmission is categorized into inductive, resonant, and microwave wireless power transmission. Magnetic Coupling Resonant Wireless Power Transfer (MCR-WPT), which uses the principle of near-field magnetic resonance to achieve wireless energy transmission, is currently the most widely studied wireless power transmission method.

[0004] Transmission power and efficiency are key research areas in magnetically coupled resonant wireless power transmission technology. Current approaches to improving these technologies are based on known load and mutual inductance. However, in practical MCR-WPT systems, load resistance and mutual inductance vary, hindering theoretical research from being applicable to real-world applications. Therefore, research on methods for identifying load and mutual inductance parameters in MCR-WPT systems is crucial for improving transmission performance in practical systems. Current approaches to identifying load and mutual inductance in MCR-WPT systems suffer from the following common issues: ① Only one parameter, either the load or the mutual inductance, can be identified; ② While changing the system operating frequency allows for load and mutual inductance identification, this approach compromises system transmission performance; ③ Adding a controller or other auxiliary modules to achieve load and mutual inductance identification effectively increases system size and design cost; and ④ High system modeling requirements and long algorithm iteration times.

[0005] To address these issues, the present invention proposes an improved Grey Wolf algorithm parameter identification method for wireless power transmission systems. This method transforms the parameter identification problem into a problem of minimizing the error between the actual measured and calculated values ​​of the transmitter's resonant current by establishing a mathematical model. This method only requires detecting the transmitter's current and the inverter circuit's output voltage, effectively reducing circuit complexity and system size, saving costs while enabling faster load and mutual inductance identification, ensuring optimal operation of the entire wireless charging system. Summary of the Invention

[0006] The technical problem to be solved by this invention is to propose an improved Grey Wolf algorithm parameter identification method for wireless power transmission systems. The main purpose of this method is to establish a mathematical model to transform the parameter identification problem into the problem of minimizing the error between the actual measured and calculated values ​​of the transmitter's resonant current. Using the improved Grey Wolf algorithm within the swarm intelligence optimization algorithm, the load and mutual inductance of a specific wireless power transmission system are identified, thereby ensuring that the entire system maintains an optimal operating environment. Furthermore, this method reduces circuit complexity and system size by only detecting the transmitter current and the inverter circuit output voltage.

[0007] The technical solution of the present invention for solving the technical problem is as follows: first, a circuit model of a wireless power transmission system is established; formulas such as the equivalent impedance, current, resonant angular frequency, and power of the entire resonant network are derived through formulas such as Kirchhoff's voltage law; finally, an efficiency expression of the entire wireless power transmission system is derived; constants and variables of the obtained efficiency expression are analyzed; theoretical analysis is performed on the obtained variables such as mutual inductance and load; and by establishing a mathematical model including the output resonant current, load, and mutual inductance of a high-order harmonic inverter circuit, the system parameter identification problem is converted into a problem related to the minimum value of the resonant current at the transmitting end; an improved gray wolf algorithm is used to optimize the set variables; and finally, the parameter identification problem of the load and mutual inductance is completed, so that the entire system operates in a relatively optimal environment.

[0008] The present invention proposes an improved Grey Wolf algorithm parameter identification method based on a wireless power transmission system.

[0009] The wireless power transmission system is composed of a transmitting series circuit and a receiving series circuit connected wirelessly;

[0010] The DC power supply, inverter circuit module, transmitter resonant compensation capacitor, transmitter coil inductor, and transmitter coil equivalent internal resistance are sequentially connected to form a transmitter series circuit;

[0011] The receiving coil inductance, receiving coil equivalent resistance, receiving coil resonant compensation capacitor, and load are connected in sequence to form a receiving series circuit;

[0012] The improved gray wolf algorithm parameter identification method comprises the following steps:

[0013] Step 1: Construct a power and efficiency model for wireless power transmission based on the inverter circuit output voltage, resonant frequency, mutual inductance, load resistance, equivalent resistance of the transmitting coil, and equivalent resistance of the receiving coil;

[0014] Step 2: According to Kirchhoff’s voltage law and superposition theorem, we can get the mutual inductance M with respect to R. LFunction expression, it is deduced that after the value of higher harmonic n is determined, the mutual inductance M and load R L The mapping relationship between the two parameters is uniquely determined.

[0015] Step 3: Combine the wireless power transmission efficiency model described in steps 1 and 2 to build a parameter identification target model. Use the improved gray wolf algorithm to transform the load and mutual inductance identification problem of the MCR-WPT system into a problem of solving the minimum error between the actual detection value and the calculated value of the resonant current at the transmitter. The load resistance and mutual inductance are optimally identified by minimizing the error function value.

[0016] Step 4: When designing the entire wireless charging system, when it comes to identifying the parameters of the load and mutual inductance, the load R L And mutual inductance M is optimized, and the load R is identified L The mutual inductance M determines the parameters of the wireless power transmission system, so that the system can operate in an optimal environment.

[0017] Preferably, the wireless power transmission power and efficiency model in step 1 is:

[0018] The system input power is:

[0019] Among them, u in is the output voltage of the inverter circuit, ω is the resonant angular frequency, M is the mutual inductance, R L is the load resistance, R1 is the equivalent internal resistance of the transmitting coil, and R2 is the equivalent internal resistance of the receiving coil.

[0020] The system output power is:

[0021] Among them, u in is the output voltage of the inverter circuit, ω is the resonant angular frequency, M is the mutual inductance, R L is the load resistance, R1 is the equivalent internal resistance of the transmitting coil, and R2 is the equivalent internal resistance of the receiving coil.

[0022] The system transmission efficiency is:

[0023] Among them, u in is the output voltage of the inverter circuit, ω is the resonant angular frequency, M is the mutual inductance, R L is the load resistance, R1 is the equivalent internal resistance of the transmitting coil, and R2 is the equivalent internal resistance of the receiving coil.

[0024] The angular frequency ω mentioned in step 1 is a constant under the resonant condition, and the other equivalent resistances and inductances can also be regarded as constants. Therefore, the mutual inductance M and the load resistance R L These are all variables to be optimized;

[0025] As a preference, the mutual inductance M in step 2 is about R L The function expression is:

[0026]

[0027] Where n is the number of higher harmonics, ω is the resonant angular frequency, M is the mutual inductance, L1 is the equivalent inductance of the transmitting coil, L2 is the equivalent inductance of the receiving coil, C1 is the resonant capacitor at the transmitting end, C2 is the resonant capacitor at the receiving end, and R L is the load resistance, R1 is the equivalent internal resistance of the transmitting coil, R2 is the equivalent internal resistance of the receiving coil, U dc is the DC power supply, I n is the peak value of the nth harmonic of the transmitter current.

[0028] Preferably, the transmitter resonant current expression at time T0 in step 3 is:

[0029]

[0030] Wherein, T0 is the starting time of the first cycle after the system runs stably, T is the system operation cycle, and T0=mT.

[0031] The mutual inductance and load parameter optimization target model constructed in step 3 is:

[0032]

[0033] Where Y(x)=i p (x)-i p (x) mea ,i p (x) mea It is the sampling value of the resonant current at the transmitter when the system is running in steady state.

[0034] Step 3 describes the target model optimization using the improved grey wolf algorithm. The specific process is as follows:

[0035] Step 3.1: Initialize the population, the number of iterations M, and the population lower bound l b , the upper bound u bGray wolves belong to the Canidae family and are considered apex predators, occupying the top of the food chain. Gray wolves tend to live in groups, with an average of 5-12 individuals per pack. They maintain a strict social hierarchy. The first level of the pyramid is the pack leader, known as α. Within the pack, α is the individual with management capabilities, primarily responsible for decisions regarding hunting, sleeping times and locations, and food distribution. The second level of the pyramid is α's think tank, known as β. β primarily assists α in making decisions. When the pack's α becomes vacant, β takes over. β's dominance within the pack is second only to α. It transmits α's orders to other pack members and provides feedback to α on how well they are executing them, acting as a bridge. The third level of the pyramid is δ, which follows the decisions of α and β and is primarily responsible for scouting, surveillance, and guarding. Poorly adapted α and β individuals will be demoted to δ. At the bottom of the pyramid is ω, which is primarily responsible for maintaining balance within the pack.

[0036] Another fascinating social behavior of gray wolves is group hunting. Social hierarchy plays a crucial role in pack hunting, with the alpha leading the hunt. Gray wolf hunting involves three main components: stalking, chasing, and approaching prey; pursuing, surrounding, and harassing prey until it stops moving; and attacking the prey.

[0037] During hunting, the behavior of gray wolves surrounding prey is defined as follows:

[0038] The distance between the individual and the prey and the position update formula of the gray wolf

[0039]

[0040] Where t is the current iteration number, and is the coefficient vector, and are the position vector of the prey and the position vector of the gray wolf respectively. and The calculation formula is as follows:

[0041]

[0042] in, is the convergence factor, which decreases linearly from 2 to 0 as the number of iterations decreases. and The modulus gets a random number between [0,1].

[0043] Gray wolves are able to identify the location of their prey and surround them. Once the wolves have identified their prey's location, β and δ, led by α, guide the pack to encircle the prey. In the decision space of the optimization problem, the optimal solution (the prey's location) is unknown. Therefore, to simulate the hunting behavior of gray wolves, we assume that α, β, and δ have better knowledge of the prey's potential location. The three best solutions obtained so far are saved and their positions are used to determine the prey's location. At the same time, the other gray wolves (including ω) are forced to update their positions based on the location of the best individual, gradually approaching the prey.

[0044] The mathematical model of how a gray wolf tracks its prey is described as follows:

[0045]

[0046] in, and Represent the distances between α, β and δ and other individuals respectively. Represent the current positions of α, β and δ respectively; is a random vector, This is the current location of the Gray Wolf.

[0047] The step length and direction of the wolf pack's individual ω moving towards α, β, and δ:

[0048]

[0049] in, and Represent the distances between α, β and δ and other individuals respectively. Represent the current positions of α, β and δ respectively; is a random vector, This is the current location of the Gray Wolf.

[0050] The final position of ω

[0051]

[0052] in, This is the current location of the Gray Wolf.

[0053] When the prey stops moving, the wolf completes the hunting process by attacking. In order to simulate approaching the prey, The value of is gradually reduced. The fluctuation range of also decreases. In other words, during the iteration process, when When the value of linearly decreases from 2 to 0, the corresponding The value of also varies in the interval [-a,a].

[0054] when When the value of is within the interval, the next position of the gray wolf can be anywhere between its current position and the position of the prey. When α, β, and δ are equal, the wolf pack attacks the prey (falling into a local optimum). The gray wolves search for prey based on the positions of α, β, and δ. The gray wolves separate from each other when searching for prey, and then gather together to attack the prey.

[0055] Based on the divergence of data modeling, we can use or The random value of is used to force the wolf to separate from the prey, which emphasizes exploration and allows the GWO algorithm to search for the optimal solution globally. The GWO algorithm has another component to help discover new solutions.

[0056] From the formula, we can see that is a random value between [0, 2]. C represents the random weight of the influence of the wolf's location on the prey. C > 1 indicates a significant influence, while C > 1 indicates a minimal influence. This helps the GWO algorithm behave more randomly and supports exploration, while also avoiding local optima during the optimization process. Furthermore, unlike A, C decreases nonlinearly, providing a global search of the decision space from the initial to the final iteration. The randomness of C plays a crucial role in avoiding local optima when the algorithm is stuck in a difficult-to-break-out local optimum, especially in the final iterations where the global optimal solution must be achieved.

[0057] Step 3.2: Improve the Gray Wolf Algorithm in the swarm intelligence optimization algorithm. Use Tent mapping to generate N relatively evenly distributed initial solutions. Then use reverse learning to generate a corresponding reverse solution for each initial solution. Compare the initial solution with the reverse solution, and select N individuals with better fitness as the initial population, resulting in a more evenly distributed initial population. The Tent mapping iteration formula is:

[0058]

[0059] Among them, b∈(0,1), X n ∈[0,1], n=1,2,...,n.

[0060] In the Gray Wolf Algorithm, when the coefficient A>1, the algorithm performs a global search; when the coefficient A<1, the algorithm performs a precise local search. A changes with the convergence factor a, so the convergence factor a of the Gray Wolf Algorithm is a key parameter for the global search and local search of the algorithm. The convergence factor of the standard Gray Wolf Algorithm decreases linearly from 2 to 0, which is difficult to adapt to the actual search process. Therefore, a convergence factor a based on the Sigmoid function is proposed, that is,

[0061]

[0062] Cauchy mutation has a stronger perturbation capability, while Gaussian mutation focuses on searching the local area near the original individual, which helps the algorithm quickly and accurately find the global minimum. Therefore, using Cauchy mutation in the early stages of the algorithm can prevent the algorithm from falling into local optimal solutions, while using Gaussian mutation in the later stages of the algorithm can perform detailed local searches and accelerate the algorithm's convergence.

[0063] Step 3.3: Use the improved Grey Wolf algorithm to transform the load and mutual inductance identification problem of the MCR-WPT system into a problem of solving the minimum error between the actual detection value and the calculated value of the resonant current at the transmitter end. The load resistance and mutual inductance values ​​are identified by minimizing the error function value.

[0064] Compared with the prior art, the present invention has the following advantages:

[0065] The present invention derives an efficiency formula for a circuit model of a compensation network SS in a wireless power transmission system and analyzes variables that affect the efficiency of the entire system.

[0066] This invention improves upon the Gray Wolf Algorithm by adding a chaotic reverse learning initialization strategy, an improved convergence factor nonlinear reduction strategy, and different application strategies for Gaussian and Cauchy mutations. By establishing a mathematical model, the parameter identification problem is transformed into a problem of minimizing the error between the actual measured and calculated values ​​of the transmitter resonant current. This method only requires detecting the transmitter current and the inverter circuit output voltage, effectively reducing circuit complexity and system size. This method effectively prevents the Gray Wolf Algorithm from falling into local optimal solutions, allowing for identification of the two parameter variables and enhancing optimization capabilities, leading to better overall design of wireless power transmission systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] Figure 1 : A flow chart of a method for identifying parameters of a wireless power transmission system using an improved grey wolf algorithm according to the present invention.

[0068] Figure 2 : Schematic diagram of the circuit of the SS compensation network wireless charging system described in the present invention.

[0069] Figure 3 : Algorithm flow chart of the improved grey wolf algorithm described in the present invention.

[0070] Figure 4 : Convergence result diagram of the improved grey wolf algorithm optimizing the target model described in the present invention.

[0071] Figure 5 : Result diagram of the improved grey wolf algorithm of the present invention for optimizing target model load identification.

[0072] Figure 6 : The result diagram of the improved grey wolf algorithm of the present invention for optimizing the mutual induction recognition of the target model.

[0073] Figure 7 : Result diagram of the improved grey wolf algorithm of the present invention optimizing the target model recognition error. DETAILED DESCRIPTION

[0074] The present invention will be described in further detail below with reference to the embodiments and drawings, but the embodiments of the present invention are not limited thereto.

[0075] The wireless power transmission system of the present invention includes: The present invention proposes an improved grey wolf algorithm parameter identification method based on the wireless power transmission system.

[0076] The wireless power transmission system is composed of a transmitting series circuit and a receiving series circuit connected wirelessly;

[0077] The DC power supply, inverter circuit module, transmitter resonant compensation capacitor, transmitter coil inductor, and transmitter coil equivalent internal resistance are sequentially connected to form a transmitter series circuit;

[0078] The receiving coil inductance, receiving coil equivalent resistance, receiving coil resonant compensation capacitor, and load are connected in sequence to form a receiving series circuit;

[0079] The system operating frequency is 85KHz, the DC power supply is 18V, the transmitting end resonant capacitor is 33nF, the transmitting end resonant inductor is 120uH, the receiving end resonant capacitor is 33Nf, and the receiving end resonant inductor is 120uH.

[0080] A parameter identification method for wireless power transmission system based on improved grey wolf algorithm. Figure 1 As shown, the specific steps include:

[0081] Step 1: Construct the power and efficiency model of wireless power transmission based on the inverter circuit output voltage, resonant frequency, mutual inductance, load resistance, equivalent resistance of the transmitting coil, and equivalent resistance of the receiving coil. The equivalent model of the wireless power transmission system is as follows: Figure 2 As shown;

[0082] The wireless power transmission power and efficiency model described in step 1 is:

[0083] The system input power is:

[0084] Among them, u in is the output voltage of the inverter circuit, ω is the resonant angular frequency, M is the mutual inductance, R L is the load resistance, R1 is the equivalent internal resistance of the transmitting coil, and R2 is the equivalent internal resistance of the receiving coil.

[0085] The system output power is:

[0086] Among them, u in is the output voltage of the inverter circuit, ω is the resonant angular frequency, M is the mutual inductance, R L is the load resistance, R1 is the equivalent internal resistance of the transmitting coil, and R2 is the equivalent internal resistance of the receiving coil.

[0087] The system transmission efficiency is:

[0088] Among them, u in is the output voltage of the inverter circuit, ω is the resonant angular frequency, M is the mutual inductance, R L is the load resistance, R1 is the equivalent internal resistance of the transmitting coil, and R2 is the equivalent internal resistance of the receiving coil.

[0089] The angular frequency ω mentioned in step 1 is a constant under the resonant condition, and the other equivalent resistances and inductances can also be regarded as constants. Therefore, the mutual inductance M and the load resistance R L These are all variables to be optimized;

[0090] Step 2: According to Kirchhoff’s voltage law and superposition theorem, we can get the mutual inductance M with respect to R. L Function expression, it is deduced that after the value of higher harmonic n is determined, the mutual inductance M and load R L The mapping relationship between the two parameters is uniquely determined.

[0091] The mutual inductance M in step 2 is about R L The function expression is:

[0092]

[0093] Where n is the number of higher harmonics, ω is the resonant angular frequency, M is the mutual inductance, L1 is the equivalent inductance of the transmitting coil, L2 is the equivalent inductance of the receiving coil, C1 is the resonant capacitor at the transmitting end, C2 is the resonant capacitor at the receiving end, and R L is the load resistance, R1 is the equivalent internal resistance of the transmitting coil, R2 is the equivalent internal resistance of the receiving coil, U dc is the DC power supply, I n is the peak value of the nth harmonic of the transmitter current.

[0094] Step 3: Combine the wireless power transmission efficiency model described in steps 1 and 2 to build a parameter identification target model. Use the improved gray wolf algorithm to transform the load and mutual inductance identification problem of the MCR-WPT system into a problem of solving the minimum error between the actual detection value and the calculated value of the resonant current at the transmitter. The load resistance and mutual inductance are optimally identified by minimizing the error function value.

[0095] The expression of the transmitter resonant current at time T0 in step 3 is:

[0096]

[0097] Wherein, T0 is the starting time of the first cycle after the system runs stably, T is the system operation cycle, and T0=mT.

[0098] The mutual inductance and load parameter optimization target model constructed in step 3 is:

[0099]

[0100] Where Y(x)=i p (x)-i p (x) mea ,i p (x) mea It is the sampling value of the resonant current at the transmitter when the system is running in steady state.

[0101] Step 3 describes the target model optimization by improving the gray wolf algorithm. Figure 3 As shown, the specific process is:

[0102] Step 3.1: Initialize the population, the number of iterations M, and the population lower bound l b , the upper bound u b Gray wolves belong to the Canidae family and are considered apex predators, occupying the top of the food chain. Gray wolves tend to live in groups, with an average of 5-12 individuals per pack. They maintain a strict social hierarchy. The first level of the pyramid is the pack leader, known as α. Within the pack, α is the individual with management capabilities, primarily responsible for decisions regarding hunting, sleeping times and locations, and food distribution. The second level of the pyramid is α's think tank, known as β. β primarily assists α in making decisions. When the pack's α becomes vacant, β takes over. β's dominance within the pack is second only to α. It transmits α's orders to other pack members and provides feedback to α on how well they are executing them, acting as a bridge. The third level of the pyramid is δ, which follows the decisions of α and β and is primarily responsible for scouting, surveillance, and guarding. Poorly adapted α and β individuals will be demoted to δ. At the bottom of the pyramid is ω, which is primarily responsible for maintaining balance within the pack.

[0103] Another fascinating social behavior of gray wolves is group hunting. Social hierarchy plays a crucial role in pack hunting, with the alpha leading the hunt. Gray wolf hunting involves three main components: stalking, chasing, and approaching prey; pursuing, surrounding, and harassing prey until it stops moving; and attacking the prey.

[0104] During hunting, the behavior of gray wolves surrounding prey is defined as follows:

[0105] The distance between the individual and the prey and the position update formula of the gray wolf

[0106]

[0107] Where t is the current iteration number, and is the coefficient vector, and are the position vector of the prey and the position vector of the gray wolf respectively. and The calculation formula is as follows:

[0108]

[0109] in, is the convergence factor, which decreases linearly from 2 to 0 as the number of iterations decreases. and The modulus gets a random number between [0,1].

[0110] Gray wolves are able to identify the location of their prey and surround them. Once the wolves have identified their prey, β and δ, led by α, guide the pack to surround the prey. In the decision space of the optimization problem, the optimal solution (the prey's location) is unknown. Therefore, to simulate the hunting behavior of gray wolves, we assume that α, β, and δ have better knowledge of the prey's potential location. We save the three best solutions obtained so far and use their positions to determine the prey's location. At the same time, we force other gray wolves (including ω) to update their positions based on the location of the best gray wolf, gradually approaching the prey.

[0111] The mathematical model of how a gray wolf tracks its prey is described as follows:

[0112]

[0113] in, and Represent the distances between α, β and δ and other individuals respectively.

[0114] Represent the current positions of α, β and δ respectively; is a random vector, This is the current location of the Gray Wolf.

[0115] The step length and direction of the wolf pack's individual ω moving towards α, β, and δ:

[0116]

[0117] in, and Represent the distances between α, β and δ and other individuals respectively. Represent the current positions of α, β and δ respectively; is a random vector, This is the current location of the Gray Wolf.

[0118] The final position of ω

[0119]

[0120] in, This is the current location of the Gray Wolf.

[0121] When the prey stops moving, the wolf completes the hunting process by attacking. In order to simulate approaching the prey, The value of is gradually reduced. The fluctuation range of also decreases. In other words, during the iteration process, when When the value of linearly decreases from 2 to 0, the corresponding The value of also varies in the interval [-a,a].

[0122] when When the value of is within the interval, the next position of the gray wolf can be anywhere between its current position and the position of the prey. When α, β, and δ are equal, the wolf pack attacks the prey (falling into a local optimum). The gray wolves search for prey based on the positions of α, β, and δ. The gray wolves separate from each other when searching for prey, and then gather together to attack the prey.

[0123] Based on the divergence of data modeling, we can use or The random value of is used to force the wolf to separate from the prey, which emphasizes exploration and allows the GWO algorithm to search for the optimal solution globally. The GWO algorithm has another component to help discover new solutions.

[0124] From the formula, we can see that is a random value between [0, 2]. C represents the random weight of the influence of the wolf's location on the prey. C > 1 indicates a significant influence, while C > 1 indicates a minimal influence. This helps the GWO algorithm behave more randomly and supports exploration, while also avoiding local optima during the optimization process. Furthermore, unlike A, C decreases nonlinearly, providing a global search of the decision space from the initial to the final iteration. The randomness of C plays a crucial role in avoiding local optima when the algorithm is stuck in a difficult-to-break-out local optimum, especially in the final iterations where the global optimal solution must be achieved.

[0125] Step 3.2: Improve the Gray Wolf Algorithm in the swarm intelligence optimization algorithm. Use Tent mapping to generate N relatively evenly distributed initial solutions. Then use reverse learning to generate a corresponding reverse solution for each initial solution. Compare the initial solution with the reverse solution, and select N individuals with better fitness as the initial population, resulting in a more evenly distributed initial population. The Tent mapping iteration formula is:

[0126]

[0127] Among them, b∈(0,1), X n ∈[0,1], n=1,2,...,n.

[0128] In the Gray Wolf Algorithm, when the coefficient A>1, the algorithm performs a global search; when the coefficient A<1, the algorithm performs a precise local search. A changes with the convergence factor a, so the convergence factor a of the Gray Wolf Algorithm is a key parameter for the global search and local search of the algorithm. The convergence factor of the standard Gray Wolf Algorithm decreases linearly from 2 to 0, which is difficult to adapt to the actual search process. Therefore, a convergence factor a based on the Sigmoid function is proposed, that is,

[0129]

[0130] Cauchy mutation has a stronger perturbation capability, while Gaussian mutation focuses on searching the local area near the original individual, which helps the algorithm quickly and accurately find the global minimum. Therefore, using Cauchy mutation in the early stages of the algorithm can prevent the algorithm from falling into local optimal solutions, while using Gaussian mutation in the later stages of the algorithm can perform detailed local searches and accelerate the algorithm's convergence.

[0131] Step 3.3: Use the improved Grey Wolf algorithm to transform the load and mutual inductance identification problem of the MCR-WPT system into a problem of solving the minimum error between the actual detection value and the calculated value of the resonant current at the transmitter end. The load resistance and mutual inductance values ​​are identified by minimizing the error function value.

[0132] Step 4: When designing the entire wireless charging system, when it comes to identifying the parameters of the load and mutual inductance, the load R L And mutual inductance M is optimized, and the load R is identified L The mutual inductance M determines the parameters of the wireless power transmission system, so that the system can operate in an optimal environment.

[0133] The convergence result diagram of the improved grey wolf algorithm for optimizing the target model is shown in the figure below: Figure 4 As shown, it can be seen that there is a good convergence effect.

[0134] The improved grey wolf algorithm of the present invention optimizes the target model load identification result as shown in the figure Figure 5 shown.

[0135] The results of the improved gray wolf algorithm for optimizing the mutual induction recognition of the target model are shown in the figure below: Figure 6 shown.

[0136] The result of optimizing the target model recognition error by the improved gray wolf algorithm of the present invention is shown in FIG. Figure 7 shown.

[0137] It can be seen from the result diagram that the present invention is in line with the expected concept and has achieved the expected tasks and results of the invention well.

[0138] Compared with the prior art, the present invention has the following advantages:

[0139] The present invention derives an efficiency formula for a circuit model of a compensation network SS in a wireless power transmission system and analyzes variables that affect the efficiency of the entire system.

[0140] This invention improves upon the Gray Wolf Algorithm by adding a chaotic reverse learning initialization strategy, an improved convergence factor nonlinear reduction strategy, and different application strategies for Gaussian and Cauchy mutations. By establishing a mathematical model, the parameter identification problem is transformed into a problem of minimizing the error between the actual measured and calculated values ​​of the transmitter resonant current. This method only requires detecting the transmitter current and the inverter circuit output voltage, effectively reducing circuit complexity and system size. This method effectively prevents the Gray Wolf Algorithm from falling into local optimal solutions, allowing for identification of the two parameter variables and enhancing optimization capabilities, leading to better overall design of wireless power transmission systems.

[0141] The above description is only a preferred embodiment of the present invention, which certainly cannot be used to limit the scope of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and changes can be made without departing from the principles of the present invention. These improvements and changes are also considered to be within the scope of protection of the present invention.

Claims

1. A parameter identification method for an improved Grey Wolf algorithm based on a wireless power transmission system, characterized by: The wireless power transmission system is composed of a transmitting series circuit and a receiving series circuit connected wirelessly; The transmitting series circuit is composed of a DC power supply, an inverter circuit module, a transmitting end resonant compensation capacitor, a transmitting coil inductor, and an equivalent internal resistance of the transmitting coil connected in sequence; The receiving series circuit is composed of a receiving coil inductor, a receiving coil equivalent resistor, a receiving coil resonant compensation capacitor, and a load connected in sequence; The improved gray wolf algorithm parameter identification method comprises the following steps: Step 1: Construct a wireless power transmission power model and a wireless power transmission efficiency model based on the inverter circuit output voltage, resonant frequency, mutual inductance, load resistance, equivalent resistance of the transmitting coil, and equivalent resistance of the receiving coil; Step 2: According to Kirchhoff’s voltage law and superposition theorem, we can get the mutual inductance M with respect to R. L Function expression, derive the value of higher harmonic order n, mutual inductance M and load R L The mapping relationship between two parameters is uniquely determined; Step 3: Build a parameter identification target model based on the wireless power transmission efficiency model. Use the improved Grey Wolf algorithm to transform the load and mutual inductance identification problem of the MCR-WPT system into a problem of minimizing the error between the actual detected value and the calculated value of the resonant current at the transmitter. The load resistance and mutual inductance are optimally identified by minimizing the error function. Step 4: When designing the entire wireless charging system, when it comes to identifying the parameters of the load and mutual inductance, the load R L And mutual inductance M is optimized, according to the load R after identification L The mutual inductance M determines the parameters of the wireless power transmission system so that the system can operate in an optimal environment; The wireless power transmission power model described in step 1 is: The system input power is: Among them, u in is the output voltage of the inverter circuit, ω is the resonant angular frequency, M is the mutual inductance, R L is the load resistance, R1 is the equivalent internal resistance of the transmitting coil, and R2 is the equivalent internal resistance of the receiving coil; The system output power is: Among them, u in is the output voltage of the inverter circuit, ω is the resonant angular frequency, M is the mutual inductance, R L is the load resistance, R1 is the equivalent internal resistance of the transmitting coil, and R2 is the equivalent internal resistance of the receiving coil; The wireless power transmission efficiency model described in step 1 is: Where ω is the resonant angular frequency, M is the mutual inductance, and R L is the load resistance, R1 is the equivalent internal resistance of the transmitting coil, and R2 is the equivalent internal resistance of the receiving coil; The angular frequency ω described in step 1 is a constant under the resonant condition, and the other equivalent resistances and inductances are considered constants; therefore, the mutual inductance M and the load resistance R L These are all variables to be optimized; The mutual inductance M in step 2 is about R L The function expression is: Where n is the number of higher harmonics, ω is the resonant angular frequency, M is the mutual inductance, L1 is the equivalent inductance of the transmitting coil, L2 is the equivalent inductance of the receiving coil, C1 is the resonant capacitor at the transmitting end, C2 is the resonant capacitor at the receiving end, and R L is the load resistance, R1 is the equivalent internal resistance of the transmitting coil, R2 is the equivalent internal resistance of the receiving coil, U dc is the DC power supply, I n is the peak value of the nth harmonic of the transmitter current; In step 3, the transmitter resonant current expression at time T0 is: Wherein, T0 is the starting time of the first cycle after the system runs stably, T is the system operation cycle, T0 = mT; In step 3, the mutual inductance and load parameter optimization target model is constructed as follows: Where Y(x)=i p (x)-i p (x) mea ,i p (x) mea It is the sampling value of the resonant current at the transmitting end when the system is in steady state operation; Step 3 describes the target model optimization using the improved grey wolf algorithm. The specific process is as follows: Step 3.1: Initialize the gray wolf population, the number of iterations M, and the population lower bound l b , the upper bound u b , dimension d; define the hierarchy of individuals in the gray wolf population; α is the group leader, responsible for group decision-making; β is responsible for assisting α in decision-making; α can take over when vacant; transmit commands and feedback; δ is responsible for obeying the orders of α and β; α or β with poor fitness may be downgraded to δ; ω is responsible for maintaining the internal balance of the population; The hunting process of gray wolf packs is completed under the leadership of α; gray wolf hunting includes the following three main parts: tracking, chasing and approaching prey; chasing, surrounding and harassing prey until it stops moving; attacking prey; During hunting, the behavior of gray wolves surrounding prey is defined as follows: The distance between the individual and the prey and the position update formula of the gray wolf Where t is the current iteration number, and is the coefficient vector, and are the position vectors of the prey and the wolf respectively; and The calculation formula is as follows: in, is the convergence factor, which decreases linearly from 2 to 0 as the number of iterations decreases, and its modulus is a random number between [0,1]; Assume that α, β, and δ know the potential location of the prey better; save the three best solutions obtained so far, and use the positions of these three to determine the location of the prey. At the same time, force other gray wolves to update their positions based on the position of the best gray wolf, gradually approaching the prey; The mathematical model of how a gray wolf tracks its prey is described as follows: in, and Represent the distances between α, β and δ and other individuals respectively; Represent the current positions of α, β and δ respectively; is a random vector, is the current position of the gray wolf; The step length and direction of the wolf pack's individual ω moving towards α, β, and δ: in, and Represent the distances between α, β and δ and other individuals respectively; Represent the current positions of α, β and δ respectively; is a random vector, is the current position of the gray wolf; The final position of ω in, is the current position of the gray wolf; The Grey Wolf Algorithm (GWO) controls the search vector by linearly decreasing the parameter a from 2 to 0. The fluctuation range [-a, a]: when Perform global exploration to avoid local optimality when When performing local development to approach the optimal solution; at the same time, introducing random vector As a perturbation weight, it provides randomness throughout the iteration process to enhance the ability to escape from local optimality; Step 3.2: Improve the gray wolf algorithm in the swarm intelligence optimization algorithm; use Tent mapping to generate n relatively evenly distributed initial solutions, and then use reverse learning to generate a corresponding reverse solution for each initial solution; compare the initial solution with the reverse solution, and select N individuals with better fitness as the initial population, obtaining a more evenly distributed initial population; the Tent mapping iteration formula is: Among them, b∈(0,1), X n ∈[0,1], n=1,2,...,n; In the Gray Wolf Algorithm, the convergence factor of the standard Gray Wolf Algorithm decreases linearly from 2 to 0, which is difficult to adapt to the actual search process; The convergence factor α of the gray wolf algorithm based on the Sigmoid function is The use of Cauchy mutation in the early stage can prevent the algorithm from falling into the local optimal solution; the use of Gaussian mutation in the later stage of the algorithm can perform detailed local search and accelerate the convergence of the algorithm; Step 3.3: Use the improved Grey Wolf algorithm to transform the load and mutual inductance identification problem of the MCR-WPT system into a problem of solving the minimum error between the actual detection value and the calculated value of the resonant current at the transmitter end. The load resistance and mutual inductance values ​​are identified by minimizing the error function value.

Images