An online prediction method for the internal temperature of the magnetic coupling mechanism of a wireless charging system based on deep learning

Abstract

This paper proposes a deep learning-based online temperature prediction method for the internal temperature of the magnetic coupling mechanism in a wireless charging system. This method uses infrared sensors to measure the surface temperature of the magnetic coupling mechanism and combines this with finite element thermal field analysis to establish a relatively accurate online temperature prediction model for the magnetic coupling mechanism. This model can be used for temperature monitoring and overtemperature protection of the magnetic coupling mechanism. Compared to traditional methods, this method can achieve real-time temperature prediction of the magnetic coupling mechanism.

Description

Technical Field

[0001] The present invention belongs to the technical field of wireless power transmission, and in particular relates to an online prediction method for the internal temperature of a magnetic coupling mechanism of a wireless charging system based on deep learning. Background Art

[0002] With the rapid development of our socioeconomic system, environmental issues are becoming increasingly severe. Increased energy hazards and carbon dioxide emissions are exacerbating the deterioration of the natural environment. The trend is for new energy vehicles, primarily electric vehicles, to replace fuel-powered vehicles. However, as electric vehicles become more widespread, the choice of charging method is crucial. Traditional electric vehicle charging methods typically involve using cables or replacing batteries, which present numerous operational challenges. For example, plugs can wear out over time, and rainy weather poses significant safety risks. Wireless charging, however, is a new type of charging technology that offers intelligent, efficient, flexible, and convenient advantages, and holds broad application prospects.

[0003] The magnetic coupling mechanism is one of the key components of the wireless charging system. When the wireless power transmission system is working, especially in high-power situations, the magnetic coupling mechanism will experience a significant temperature rise. Excessive temperature will lead to increased heat loss in the system and even cause system failure, resulting in safety hazards. Currently, temperature sensors are generally used to measure the temperature of the coupling mechanism, but metal temperature sensors are affected by high-frequency magnetic fields and cannot work normally. Infrared temperature sensors can only measure the surface temperature of the magnetic coupling and cannot determine the internal temperature of the magnetic coupling mechanism, which can easily cause safety hazards.

[0004] At present, the main research results at home and abroad are:

[0005] The University of Auckland in New Zealand studied the thermal characteristics of embedded DD-type transmitting coils under different ambient temperatures. Finite element analysis was used to predict the thermal equilibrium steady-state temperature reached by the transmitting end, and experimental verification was conducted. The measured temperature values ​​were in good agreement with the temperature values ​​obtained using the finite element analysis method.

[0006] The University of Cambridge in the UK conducted in-depth research on the heat dissipation problem of PCB boards in wireless charging systems, established a thermal resistance analytical model for PCB through-holes and pads, and proposed an algorithm for quickly obtaining the ambient thermal resistance of PCB boards and predicting semiconductor junction temperatures. The proposed thermal model and design optimization algorithm were verified through fluid dynamics simulation and experimental measurements.

[0007] The Beijing Institute of Technology research team calculated the thermal resistance of the material and established a thermal network for the heating element. They then applied power loss to the nodes as a heat source. The calculated hotspot temperatures were close to those calculated using the finite element method.

[0008] The current research results mainly have the following technical problems:

[0009] 1. Existing infrared temperature measurement technology can only measure the temperature on the surface of the magnetic coupling mechanism, but cannot measure the temperature inside the magnetic coupling mechanism;

[0010] 2. Existing metal temperature sensors cannot work properly due to the influence of high-frequency magnetic fields and can only perform power-off temperature measurement. Summary of the Invention

[0011] This invention aims to address the challenges of the prior art by proposing a deep learning-based online temperature prediction method for the internal temperature of the magnetic coupling mechanism of a wireless charging system. This method uses infrared sensors to measure the surface temperature of the magnetic coupling mechanism and, combined with finite element thermal field analysis, establishes a relatively accurate online temperature prediction model for the magnetic coupling mechanism. This model can be used for temperature monitoring and overtemperature protection of the magnetic coupling mechanism.

[0012] The present invention is achieved through the following technical solutions. The present invention proposes an online prediction method for the internal temperature of the magnetic coupling mechanism of a wireless charging system based on deep learning. The method is specifically as follows:

[0013] Step 1: First, perform offline preparation and establish a finite element simulation model of the magnetic coupling mechanism. Build a temperature dataset by changing the coil excitation and adjusting the spatial environment.

[0014] Step 2: Based on the simulation data, a regression prediction model from the surface temperature of the magnetic coupling mechanism to the maximum temperature of the internal components is established through a deep belief network;

[0015] Step 3: Store the prediction model in the upper or lower computer for temperature prediction;

[0016] Step 4: Next, conduct online prediction. During the actual operation of the wireless charging system, use an infrared temperature sensor to measure the surface temperature of the magnetic coupling mechanism and obtain a surface temperature data matrix.

[0017] Step 5: Input the surface temperature data matrix into the prediction model, and complete the online prediction of the internal temperature of the magnetic coupling mechanism through the core processor of the upper or lower computer;

[0018] Step 6: Output the online prediction results of the internal temperature of the magnetic coupling mechanism.

[0019] The beneficial effects of the present invention are:

[0020] 1. Compared with traditional methods, the method of the present invention can realize real-time prediction of the internal temperature of the magnetic coupling mechanism.

[0021] 2. This method does not require destructive dismantling of the magnetic coupling mechanism, and does not require setting test points inside the magnetic coupling mechanism. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 Schematic diagram of the online prediction method for the internal temperature of the magnetic coupling mechanism. DETAILED DESCRIPTION

[0023] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0024] The present invention discloses an online prediction method for the internal temperature of a magnetic coupling mechanism of a wireless charging system. Figure 1 As shown:

[0025] First, a finite element simulation model is established for the magnetic coupling mechanism. Then, a regression prediction model from the surface temperature of the magnetic coupling mechanism to the maximum temperature of the internal components is established based on the simulation data through the relevant deep learning algorithm. The data model is stored in the core processor for temperature prediction.

[0026] Then, during the actual operation of the wireless charging system, an infrared temperature sensor is used to measure the surface temperature of the magnetic coupling mechanism to obtain a surface temperature data matrix. The data is then input into the prediction model, and the online prediction of the internal temperature of the magnetic coupling mechanism is completed through the model calculation of the core processor.

[0027] Example

[0028] The present invention proposes a method for online prediction of the internal temperature of the magnetic coupling mechanism of a wireless charging system based on deep learning. The method is specifically as follows:

[0029] Step 1: First, perform offline preparation and establish a finite element simulation model of the magnetic coupling mechanism. Build a temperature dataset by changing the coil excitation and adjusting the spatial environment.

[0030] Specifically, offline preparations were first performed. Based on the actual structural parameters of the system, a finite element thermal magnetic field simulation model of the magnetic coupling mechanism was established using ANSYS Maxwell. A magnetic field loss analysis was performed first, and then the loss data was substituted into the thermal field to simulate and solve the temperature conditions of the system under different operating conditions. A temperature data set was constructed by changing the coil excitation and adjusting the spatial environment.

[0031] Step 2: Based on the simulation data, a regression prediction model from the surface temperature of the magnetic coupling mechanism to the maximum temperature of the internal components is established through a deep belief network;

[0032] Specifically, a regression prediction model for the temperature of the magnetic coupling mechanism surface to the maximum temperature of the internal components is established using a deep belief network training method based on simulation data. First, forward stacking learning and reverse fine-tuning learning are performed on several restricted Boltzmann machines in the deep belief network. After all restricted Boltzmann machines in the deep belief network are trained, reverse fine-tuning learning is used using simulation labeled data to train and update the parameters of the deep belief network model, ultimately obtaining a relatively accurate temperature regression prediction model.

[0033] Step 3: Store the prediction model in the upper or lower computer for temperature prediction;

[0034] Step 4: Next, conduct online prediction. During the actual operation of the wireless charging system, use an infrared temperature sensor to measure the surface temperature of the magnetic coupling mechanism and obtain a surface temperature data matrix.

[0035] Step 5: Input the surface temperature data matrix into the prediction model, and complete the online prediction of the internal temperature of the magnetic coupling mechanism through the core processor of the upper or lower computer;

[0036] Step 6: Output the online prediction results of the internal temperature of the magnetic coupling mechanism.

Claims

1. A method for online prediction of the internal temperature of the magnetic coupling mechanism of a wireless charging system based on deep learning, characterized by: The method is specifically as follows: Step 1: First, perform offline preparations and establish a finite element thermal magnetic field simulation model of the magnetic coupling mechanism. Perform magnetic field loss analysis first, then substitute the loss data into the thermal field to simulate and solve the temperature conditions of the system under different working conditions. Construct a temperature data set by changing the coil excitation and adjusting the spatial environment. Step 2: Based on simulation data, a regression prediction model is established using a deep belief network to derive the temperature from the surface temperature of the magnetic coupling mechanism to the maximum temperature of the internal components. First, forward stacking learning and reverse fine-tuning are performed on several restricted Boltzmann machines in the deep belief network. After all restricted Boltzmann machines in the deep belief network are trained, reverse fine-tuning learning is used to train and update the parameters of the deep belief network model using simulation labeled data, ultimately obtaining a relatively accurate temperature regression prediction model. Step 3: Store the prediction model in the upper or lower computer for temperature prediction; Step 4: Next, conduct online prediction. During the actual operation of the wireless charging system, use an infrared temperature sensor to measure the surface temperature of the magnetic coupling mechanism and obtain a surface temperature data matrix. Step 5: Input the surface temperature data matrix into the prediction model, and complete the online prediction of the internal temperature of the magnetic coupling mechanism through the core processor of the upper or lower computer; Step 6: Output the online prediction results of the internal temperature of the magnetic coupling mechanism.

Images