CN-121981165-A - Capacitance parameter compensation system for multichannel wireless power supply system

CN121981165ACN 121981165 ACN121981165 ACN 121981165ACN-121981165-A

Abstract

The invention discloses a capacitance parameter compensation system for a multichannel wireless power supply system, which relates to the technical field of wireless power transmission and comprises a state sensing module, an intelligent decision module and a deep reinforcement learning intelligent agent, wherein the state sensing module is used for acquiring voltage, current, phase and load demand information of each channel in the multichannel wireless power supply system in real time and generating a state vector describing the real-time running state of the system, and the intelligent decision module is deployed with the deep reinforcement learning intelligent agent and is used for receiving the state vector and calculating according to a pre-trained strategy network. The capacitance parameter compensation system provided by the invention is aimed at a capacitance parameter compensation system provided by a multichannel wireless power supply system, so that the flexibility and the efficiency of the system operation are obviously improved, and the accurate sensing of the system operation state is realized by collecting key parameters such as voltage and current of each channel in real time and constructing a multidimensional state vector, thereby introducing a deep reinforcement learning intelligent body and effectively aiming at complex working conditions such as load demand change, inter-channel interference and the like.

Inventors

LIANG YAN
WANG RONG
YAN CHUNPING
XU LUNING
ZHAO YING

Assignees

江西职业技术大学

Dates

Publication Date: 20260505
Application Date: 20251230

Claims (10)

1. A capacitance parameter compensation system for a multi-channel wireless power supply system, comprising: the state sensing module is used for collecting voltage, current, phase and load demand information of each channel in the multichannel wireless power supply system in real time and generating a state vector describing the real-time running state of the system; the intelligent decision module is deployed with a deep reinforcement learning intelligent agent and is used for receiving the state vector, calculating according to a pre-trained strategy network and outputting an optimal capacitance coding instruction corresponding to each channel compensation capacitor; The execution module comprises a multipath digital adjustable capacitor array and is used for receiving the capacitance coding instruction and adjusting the capacitance value in each channel resonance compensation network so as to dynamically optimize the transmission performance of the system; And the training and fine tuning module is used for carrying out strategy training and online safety fine tuning on the deep reinforcement learning intelligent body through a defined reward function in a simulation environment or an actual system so as to adapt to system parameter changes.
2. The capacitive parameter compensation system for a multi-channel wireless power supply system of claim 1, wherein the state sensing module specifically comprises: Based on the main circuit topology of the multichannel wireless power supply system, high-precision voltage and current sensors are deployed at each compensation node of the primary side and the secondary side of the system, and real-time voltage and current sampling data of each channel are obtained; Based on the real-time voltage and current sampling data, the amplitude and phase difference of the voltage and current fundamental wave under the working frequency of each channel are calculated through a phase-locked loop or a digital signal processing unit, and the real-time impedance phase angle of each channel is obtained; Based on wireless communication or power line carrier, receiving real-time load demand or output power instruction fed back from each receiving end; And integrating the voltage amplitude, the current amplitude, the impedance phase angle and the load demand of each channel to construct a multidimensional state vector for describing the global running state of the system.
3. The capacitance parameter compensation system for a multi-channel wireless power supply system according to claim 2, wherein the intelligent decision module specifically comprises: Deploying an intelligent agent based on a deep reinforcement learning algorithm, wherein a strategy network and a value network of the intelligent agent are formed by a deep neural network; the input of the strategy network is a multidimensional state vector provided by a state sensing module, and the output is a capacitance coding motion vector for controlling the digital adjustable capacitor array of each channel; The value network is used to evaluate long-term expected returns obtained from taking a particular action in a given state; A reward function is designed which calculates a weighted sum based on the total transmission efficiency of the system at the next moment, the mean square error of the output power of each channel and the target value, and the total harmonic distortion of the output voltage of the system.
4. A capacitance parameter compensation system for a multi-channel wireless power supply system according to claim 3 and wherein said intelligent decision module further comprises: In the off-line deployment stage of the system, a simulation model of a multichannel wireless power supply system based on high fidelity is utilized to pretrain a deep reinforcement learning intelligent agent in a large scale by utilizing a near-end strategy optimization algorithm, so as to generate an initial strategy model; in the online operation stage of the system, loading a pre-trained initial strategy model to an embedded processor, receiving a state vector in a fixed period and reasoning and outputting an action vector in real time; Setting a safety exploration boundary and limiting the adjustment step length and range of the capacitance value in the motion vector.
5. The capacitive parameter compensation system for a multi-channel wireless power supply system of claim 4, wherein said execution module specifically comprises: The digital adjustable capacitance units formed by binary capacitance arrays and switch networks are connected in parallel or in series at the resonance compensation network nodes of the channels; the binary capacitor array is formed by connecting a plurality of fixed capacitors with binary values in series with high-speed MOSFET switches, and discrete digital adjustment of the capacitance values is realized through switch combination; and receiving the capacitance coding motion vector from the intelligent decision module, driving each channel switching network through the high-speed digital isolation interface, and converting the target capacitance code into a specific switching control signal.
6. The capacitive parameter compensation system for a multi-channel wireless power supply system of claim 5, wherein said execution module further comprises: the front end of the digital adjustable capacitor unit is provided with a protection circuit which comprises an overvoltage protection device and a buffer circuit so as to prevent the voltage impact under high-frequency resonance from damaging a switching device; And adopting a field programmable gate array to carry out synchronous time sequence management on the multi-path switch control signals.
7. The capacitive parameter compensation system for a multi-channel wireless power supply system of claim 6, wherein the training and fine tuning module comprises: Constructing a system simulation environment comprising a multi-channel cross mutual inductance matrix, device parasitic parameters and a load dynamic change model, and taking the system simulation environment as an offline training field of an intelligent body; In the simulation environment, updating the neural network parameters through the time steps of interaction between the intelligent agent and the simulation environment until the strategy is converged; the state-action-rewards data pairs collected in actual operation are stored to an experience playback buffer periodically for subsequent online policy fine tuning.
8. The capacitive parameter compensation system for a multi-channel wireless power supply system of claim 7, wherein said training and fine tuning module further comprises: Setting an on-line fine tuning trigger, and triggering a fine tuning process when the system detects that the rewarding function value is lower than a set threshold value or the load is subjected to step change; In the fine tuning process, sampling data from an experience playback buffer area, and performing incremental training on a strategy network of an intelligent agent at a learning rate; The trimming process is performed asynchronously on the coprocessor.
9. The capacitive parameter compensation system for a multi-channel wireless power supply system of claim 8, further comprising a system co-operation flow: the state sensing module collects data according to a fixed control period and generates a state vector; The intelligent decision module completes the receiving of the state vector, the neural network reasoning and the output of the motion vector in each control period; The execution module completes the adjustment of the digital adjustable capacitance values of all channels before the next control period starts; and the training and fine tuning module runs asynchronously in the background of the system to finish the updating of the model and the strategy optimization.
10. The capacitance parameter compensation system for a multi-channel wireless power supply system according to claim 9, wherein the dynamic optimization target executed by the intelligent decision module is a global performance index of the system, and the index is a multi-target weighting function of the total transmission efficiency of the system, the output power balance degree of each channel and the working stability of the system; The total transmission efficiency of the system is calculated according to the total input power of the primary side and the total output power of all the secondary sides; The output power balance degree of each channel is represented by calculating the root mean square value of the difference between the actual output power and the target power of each channel; The working stability of the system is evaluated by monitoring fluctuation variance of output voltage of each channel in a time window; and the deep reinforcement learning agent outputs a capacity value coding instruction by optimizing the multi-objective weighting function.

Description

Capacitance parameter compensation system for multichannel wireless power supply system Technical Field The invention relates to the technical field of wireless power transmission, in particular to a capacitance parameter compensation system for a multichannel wireless power supply system. Background The multichannel wireless power supply system can supply power to a plurality of scattered loads simultaneously, is widely applied to the fields of internet of things terminals, industrial automation equipment, new energy storage devices and the like, and becomes an important development direction of wireless power transmission technology. The transmission performance of the system is highly dependent on the parameter matching of the resonance compensation network, wherein the capacitance parameter is used as a core component of the resonance loop, and directly influences the transmission efficiency, the output power precision and the interference suppression effect among all channels of the system. In the actual operation process, the multichannel wireless power supply system faces complex working conditions such as dynamic change of load demands, cross mutual inductance interference among channels, parasitic parameter drift of devices and the like, so that strict requirements are put forward on the real-time dynamic compensation of capacitance parameters, and a reasonable capacitance parameter compensation scheme becomes a key for guaranteeing stable and efficient operation of the system. Most of the existing capacitance parameter compensation schemes are based on preset working conditions for fixed parameter design, or a simple closed-loop adjustment strategy is adopted, so that complex dynamic characteristics of a multichannel system are difficult to comprehensively adapt. Although the partial dynamic compensation scheme can realize parameter adjustment, the problems of insufficient response speed and limited adjustment precision exist, and the multi-objective requirements of the total transmission efficiency of the system, the power balance degree of each channel and the working stability cannot be met by focusing on single performance index optimization. Meanwhile, the existing scheme lacks an adaptive intelligent decision mechanism in the face of coupling interference among multiple channels and parameter time-varying characteristics, and is difficult to quickly generate an optimal compensation strategy, so that the system is easy to generate resonance offset, transmission performance fluctuation and other conditions under complex working conditions. In this regard, we propose a capacitance parameter compensation system for a multi-channel wireless power supply system. Disclosure of Invention In order to solve the technical problems, the capacitive parameter compensation system for the multichannel wireless power supply system is provided, the capacitive parameter compensation scheme in the prior art is designed for fixed parameters or is simple closed-loop adjustment, complex dynamic characteristics of the multichannel system are difficult to adapt, and response speed and adjustment accuracy are limited. The method has the advantages that the optimization of single performance indexes is emphasized in multiple ways, an adaptive intelligent decision mechanism is lacked, and the problems of system resonance offset and transmission performance fluctuation are easily caused due to difficulty in coping with channel coupling interference and parameter time-varying. In order to achieve the above purpose, the invention adopts the following technical scheme: A capacitance parameter compensation system for a multi-channel wireless power supply system, comprising: the state sensing module is used for collecting voltage, current, phase and load demand information of each channel in the multichannel wireless power supply system in real time and generating a state vector describing the real-time running state of the system; the intelligent decision module is deployed with a deep reinforcement learning intelligent agent and is used for receiving the state vector, calculating according to a pre-trained strategy network and outputting an optimal capacitance coding instruction corresponding to each channel compensation capacitor; The execution module comprises a multipath digital adjustable capacitor array and is used for receiving the capacitance coding instruction and adjusting the capacitance value in each channel resonance compensation network so as to dynamically optimize the transmission performance of the system; And the training and fine tuning module is used for carrying out strategy training and online safety fine tuning on the deep reinforcement learning intelligent body through a defined reward function in a simulation environment or an actual system so as to adapt to system parameter changes. Preferably, the state sensing module specifically includes: Based on the main circuit topology of the multichannel wi