Search

CN-115758866-B - Transmission performance optimization method for wireless electric energy transmission system of electric automobile

CN115758866BCN 115758866 BCN115758866 BCN 115758866BCN-115758866-B

Abstract

The invention is applicable to the technical field of wireless power transmission, and provides a transmission performance optimization method of an electric vehicle wireless power transmission system, which comprises the following steps of 1, quantifying uncertainty of transmission efficiency of an electric vehicle WPT system; and step 2, improving a multi-objective optimization algorithm according to the result in the step 1, and step 3, optimizing the transmission performance of the WPT system of the electric automobile through the improved algorithm in the step 2. The improved multi-objective hawk optimization algorithm is utilized to carry out optimization design on the WPT system compensation circuit and the magnetic energy coil group structure, so that the charging efficiency and the robustness of the WPT system of the electric automobile can be synchronously improved, the solving precision and the computing efficiency are improved by utilizing a mode of integrating and learning through a plurality of networks, the original optimization algorithm is improved, and the searching precision and the searching efficiency of the algorithm on the optimal structural parameter group of the WPT system are effectively improved.

Inventors

  • WANG TIANHAO
  • WU YANGYUN
  • LI BO
  • YU QUANYI
  • XU LINLIN
  • GAO LE

Assignees

  • 吉林大学

Dates

Publication Date
20260512
Application Date
20221102

Claims (7)

  1. 1. The transmission performance optimization method of the wireless electric energy transmission system of the electric automobile is characterized by comprising the following steps of: Step 1, quantifying uncertainty of transmission efficiency of a WPT system of an electric automobile; Step 2, improving a multi-objective optimization algorithm according to the result in the step 1; Step 3, optimizing the transmission performance of the WPT system of the electric automobile through the improved algorithm in the step 2; the quantifying the uncertainty of the transmission efficiency of the WPT system of the electric automobile comprises the following steps: s1, selecting a plurality of random variables influencing transmission efficiency by combining possible conditions of an electric automobile WPT system in an actual charging process, and determining the distribution type of each variable; S2, establishing a deep learning neural network model; the improvement of the multi-objective optimization algorithm comprises the following steps: S3, sampling data from the distribution range of the random variables in the S1 to serve as training samples of the network model, wherein transmission efficiency values corresponding to each group of samples serve as training labels, and the model is trained; s4, sampling quantitative data in the distribution range of the variables to serve as a test sample of a model, and calculating to obtain the mean value, variance and probability density distribution function statistical characteristic parameters of the wireless power transmission efficiency of the WPT system; S5, fusing the Tent chaotic mapping and the self-adaptive inertia weight strategy to a multi-objective hawk optimization algorithm to obtain an improved multi-objective hawk optimization algorithm; the method for optimizing the transmission performance of the WPT system of the electric automobile comprises the following steps: S6, determining structural parameters of the WPT system to be optimized; S7, optimizing the WPT system structure by taking the mean value and the variance of the transmission efficiency in the fluctuation range as optimization targets based on an improved multi-target hawk optimization algorithm; and S8, calculating by using a deep learning network to obtain a probability density distribution function of the transmission efficiency of the WPT system after optimization.
  2. 2. The method for optimizing transmission performance of the wireless power transmission system of the electric automobile according to claim 1, wherein the relation between the input quantity x and the output quantity y of a single node on an hidden layer in the deep learning neural network model in S2 is: (1) in the formula (1), sigma () represents a nonlinear transfer function, w is a linear mapping, and b is a bias term; When the error between the output value and the training label tends to infinity, the evaluation index function of the training process is as follows: (2) In the formula (2), N represents the number of training samples, y i represents a network output value, and y represents a training label; In the data acquisition stage of the deep learning neural network model, an inclination angle alpha between a transmitting coil and a receiving coil of the WPT system of the electric automobile, a vertical distance d between the transmitting coil and the receiving coil, horizontal offsets delta x and delta y between centers of the transmitting coil and the receiving coil, a loop equivalent resistance R 1 of a transmitting end of a compensation circuit, a loop equivalent resistance R 2 of a receiving end of the compensation circuit and a load resistance R L are used as input variables of the model, and a plurality of network integration modes are adopted to build a first part DNN model, a second part DNN model and a third part DNN model to realize uncertainty quantification of the transmission efficiency of the WPT system.
  3. 3. The method for optimizing transmission performance of wireless power transmission system of electric automobile according to claim 2, wherein the input quantity of the first part DNN model is α, d, Δx, Δy, the output quantity is coil mutual inductance M, the first part DNN model is composed of six common full-connection layers and one batch normalization layer, the number of nodes from front to back of the six full-connection layers is 4, 64, 32, 16, 1, the input characteristic number of the batch normalization layer is 32, and the mathematical model is: (3) (4) (5) In the formula (3), gamma and beta are parameter vectors, default values are 0 and 1 respectively, epsilon is used for guaranteeing the stability of the numerical value, and default value is 1e -5 .
  4. 4. The method for optimizing transmission performance of wireless power transmission system of electric automobile according to claim 2, wherein the second part DNN model is a single full-connection layer and is composed of three neuron nodes, and is used for compressing the data of R 1 、R 2 、R L and improving the learning and information drawing capabilities of the model.
  5. 5. The method for optimizing transmission performance of wireless power transmission system of electric automobile according to claim 2, wherein the third part DNN model consists of five full-connection layers and dropout layers, and the number of nodes from front to back of the five full-connection layers is 4, 64, 32, 16,1 respectively; The calculation formula of the neuron node after combining with the dropout module is updated as follows: (6) (7) in the formula (6), r l represents a random number conforming to Bernoulli distribution, p represents a corresponding probability, and an ideal value of p is 0.5; After the deep learning neural network is trained, uncertainty statistical characteristic parameters of the wireless power transmission efficiency of the WPT system can be calculated by collecting quantitative data again from the distribution range of random variables of R 1 、R 2 、R L , alpha, d, delta x and delta y to serve as test samples of the deep learning neural network model.
  6. 6. The method for optimizing transmission performance of the wireless power transmission system of the electric automobile according to claim 1, wherein the optimizing process of the improved multi-objective hawk optimizing algorithm comprises a population initialization stage, a global exploration stage, a local exploration stage, a global exploitation stage and a local exploitation stage; the mathematical model for initializing the population is as follows: (8) In formula (8), X min represents the lower bound of the variable, X max represents the upper bound of the variable, and rand is a random number between 0 and 1; the mathematical model of global exploration is: (9) In the formula (9), X 1 (t+1) represents a solution generated after (t+1) iterations in the global exploration process, X best (T) represents an optimal solution obtained before the T iteration, T represents the maximum iteration number, X M (T) represents an average value of the current solution at the T iteration, and rand is a random number between 0 and 1; The mathematical model of the local exploration is as follows: (10) (11) (12) In the formulas (10), (11) and (12), X 2 (t+1) represents a solution generated after (t+1) iterations in the local exploration process, X best (t) represents an optimal solution obtained before the t iteration, X R (t) is a random solution obtained in the population scale range at the time of the t iteration, levy (D) represents a Levy flight distribution function of hawk, s and q in a mathematical model are fixed constants, values of 0.01, 1.5 are respectively taken, and μ and v represent gaussian distribution random numbers respectively obeying N (0, sigma 2 ) and N (0, 1); The mathematical model of global exploitation is: (13) In the formula (13), X 3 (t+1) represents a solution generated after (t+1) iterations in the global exploitation process, X best (t) represents an approximate position of a prey before the t iteration, X M (t) represents an average value of the current solution at the t iteration, and both zeta and delta are exploitation adjustment parameters; the mathematical model of the local exploitation is as follows: (14) (15) Wherein X 4 (t+1) represents a solution generated after (t+1) times of iteration in the local exploitation process, X (T) represents a current solution at the T-th time of iteration, T and T respectively represent the current iteration times and the maximum iteration times, g (rand) is a random number in a (-0.2,1) interval, various motion trajectories when the hawk closely tracks the prey are reflected, and a scaling factor 2 (1- (T/T)) of Levy (D) reflects the flying rate of the hawk.
  7. 7. The method for optimizing transmission performance of wireless power transmission system of electric vehicle according to claim 6, wherein in the population initialization stage, In the initial stage of AO, tent chaotic mapping is introduced, and the mathematical model of the chaotic mapping can be expressed as: (16) In the global exploration phase described above, When AO is in global exploration phase, introducing adaptive inertia weights, new AO behavior patterns can be expressed as: (17)。

Description

Transmission performance optimization method for wireless electric energy transmission system of electric automobile Technical Field The invention belongs to the technical field of wireless power transmission, and particularly relates to a transmission performance optimization method of an electric automobile wireless power transmission system. Background With the continuous acceleration of the development process of new energy electric vehicles, wireless Power Transmission (WPT) technology has become a focus of attention of many research institutions and automobile enterprises worldwide, and compared with traditional wired charging, WPT technology has the advantages of water resistance, dust resistance, safe operation, no mechanical abrasion and the like, and has quite wide application prospects. However, while the application of WPT technology for electric vehicles is becoming more popular, challenges are presented gradually, and the transmission efficiency of WPT systems is a typical problem that is sensitive to external interference factors, which limits the further development of the technology to a certain extent. In the problems, the uncertainty quantization method has important significance, and can acquire relevant statistical characteristic parameters of the transmission efficiency of the WPT system, so that guidance is further provided for structural optimization design of the WPT system. Aiming at the problem of uncertainty of the transmission efficiency of the WPT system of the electric automobile, the calculation efficiency is lower due to the fact that the uncertainty quantization is directly carried out on the expansion of the transmission efficiency by the traditional Monte Carlo method. In the current optimization research of the transmission performance of the WPT system, a single-target optimization algorithm is generally adopted to optimally design the structural parameters of the WPT system, and as the optimization target of the single-target optimization algorithm is a single object, the transmission performance of the WPT system cannot be truly improved, and therefore a multi-target optimization algorithm is needed. Disclosure of Invention The embodiment of the invention aims to provide a transmission performance optimization method of an electric automobile wireless power transmission system, which aims to solve the problem that the transmission performance of a WPT system is not ideal under the influence of uncertain factors, and the calculation efficiency is lower due to the fact that the traditional Monte Carlo method is directly adopted to quantify the uncertainty of the expansion of the transmission efficiency. In the current optimization research of the transmission performance of the WPT system, a single-target optimization algorithm is generally adopted to optimally design the structural parameters of the WPT system, and as the optimization target of the single-target optimization algorithm is a single object, the transmission performance of the WPT system cannot be truly improved, and therefore a multi-target optimization algorithm is needed. The embodiment of the invention is realized in such a way that the transmission performance of the wireless electric energy transmission system of the electric automobile is optimized, and the transmission performance optimization method of the wireless electric energy transmission system of the electric automobile comprises the following steps: step 1, quantifying uncertainty of transmission efficiency of an electric automobile WPT system: Combining the possible conditions in the wireless power transmission process of the electric automobile, taking the inclination angle between the transmitting coil and the receiving coil, the vertical distance between the transmitting coil and the receiving coil, the horizontal offset between the transmitting coil and the center of the receiving coil, the equivalent resistance of a transmitting end loop of a compensation circuit, the equivalent resistance of a receiving end loop of the compensation circuit and the load resistance as random uncertain input variables, establishing a WPT system transmission efficiency uncertainty quantization model based on a deep learning network, obtaining relevant statistical characteristic parameters such as mean value, variance, probability density distribution and the like of the transmission efficiency in a fluctuation range, and verifying the accuracy and efficiency of the uncertainty of the wireless power transmission efficiency of the deep learning quantized automobile by comparing the calculation results of a classical Monte Carlo method. Step 2, improving a multi-objective optimization algorithm: The multi-objective hawk optimization (AO) algorithm is improved, a good population distribution mechanism is formed by combining Tent chaotic mapping in the population initialization stage of the algorithm, and the exploration mode is improved by combining self-ad