Disclosure of Invention
The application provides a charging pile double-frequency coupling wireless transmission device and an efficiency compensation method, and aims to solve the problem that the existing wireless charging technology is insufficient in transmission efficiency and stability under complex working conditions.
In a first aspect, a charging pile dual-frequency coupling wireless transmission device includes:
the transmitting end comprises a low-frequency channel and a high-frequency channel, the low-frequency channel is driven by a low-frequency resonant circuit to drive a first coil (L1) and has strong magnetic field penetration characteristics, the high-frequency channel is driven by a high-frequency resonant circuit to drive a second coil (L2) so as to realize local enhanced transmission, and the switching circuit is an electronic switch array and dynamically switches a low-frequency/high-frequency excitation mode;
The receiving end comprises a receiving coil (L3) and a cooperative coil (L4) which are tightly coupled, wherein the L4 is tightly wound on the outer side of the L3, the cooperative coil (L4) and the transmitting end (L1/L2) form strong coupling, the coupling coefficient k of the L3 and the L4 is more than or equal to 0.4, and the magnetic core adopts a concave-convex magnetic core layered structure;
the composite magnetic shielding layer reduces the magnetic leakage intensity and electromagnetic radiation;
The control module is used for collecting transmission efficiency, load impedance and offset signals in real time, a double-frequency dynamic compensation algorithm is built in, and resonance capacitance (C1/C2) and switching frequency are adjusted according to efficiency feedback, so that the transmission efficiency always works at an optimal coupling point.
Further, the coil structure of the low-frequency channel is that the first coil (L1) adopts a rectangular close-wound coil structure, and the outer diameter of the first coil is 1.5-2 times of that of the receiving end coil;
The second coil (L2) adopts a grouping series wound coil structure, the turns ratio of the inner ring and the outer ring is 1:3, and the coils are filled with high dielectric constant media at intervals.
Furthermore, the magnetic shielding layer is made of PC95 manganese zinc ferrite and copper foil composite shielding material, the thickness of the shielding layer is 0.3-0.5mm, and the back and the side edges of the receiving end are covered.
Further, the working steps of the control module include:
the data acquisition, wherein a control module monitors data of offset delta d, temperature T and load current I_L in real time through a sensor and an acquisition circuit;
Step two, data processing, namely, after preprocessing the acquired data, sending the preprocessed data into a processor for further processing, and analyzing and calculating the data by the processor to obtain the current state and working parameters of the system;
and thirdly, inputting working parameters into a dynamic compensation algorithm, namely adjusting the resonant capacitor (C1/C2) and the switching frequency according to the transmission efficiency monitored in real time, so that the system always works at an optimal coupling point.
In a second aspect, a method for compensating efficiency of a charging pile includes the steps of:
s1, dual-frequency cooperative transmission, namely, initial energy transmission of a low-frequency channel, namely, when a system is started, the low-frequency channel is activated by default, and a basic energy transmission link is established by utilizing the strong magnetic field penetration characteristic of the low-frequency channel;
s2, dynamic load detection, namely monitoring the load power requirement of a receiving end in real time, and triggering a high-frequency channel to cooperatively work if the load power requirement increases suddenly;
S3, starting an offset compensation mechanism, namely detecting the offset delta d of a receiving end through a Hall sensor, and gradually increasing the power ratio of a high-frequency channel when delta d is more than 5 cm to form a composite magnetic field of 'low-frequency global coverage+high-frequency local enhancement';
S4, high-frequency channel self-adaptive power distribution, namely dynamically adjusting the ratio of high-frequency power to low-frequency power based on real-time feedback of a coupling coefficient k;
s5, pulse width modulation optimization, namely adopting a variable duty ratio PWM to control a high-frequency inverter, and increasing the high-frequency pulse width to compensate efficiency loss when shifting;
S6, dynamic parameter adjustment, namely calculating an optimal frequency ratio in real time based on load change, and adjusting the duty ratio of the inverter through a PID controller;
and S7, predicting the attenuation trend of the coupling coefficient k by using a neural network, adjusting the compensation capacitance value in advance, and inhibiting the efficiency fluctuation.
Further, the formula of the ratio of the high frequency power to the low frequency power is:
Wherein, Phigh is the allocated power of the high frequency band, Plow is the allocated power of the low frequency band, khigh is the coupling coefficient of the high frequency band, klow is the coupling coefficient of the low frequency band, RL is the return loss, Rloss is the allocated loss, and α is the environmental correction factor.
Further, the frequency ratio formula is:
Wherein fhigh is a high-frequency resonance frequency, flow is a low-frequency resonance frequency, k is a coupling coefficient, L is an inductance value, RL is return loss, and Rloss is distribution loss.
Further, the neural network predicting the attenuation trend of the coupling coefficient k in S7 includes the following steps:
s7.1, collecting and preprocessing data of parameters such as a historical coupling coefficient k, an offset delta d, a temperature T, a load current I_L and the like, cleaning the collected data, removing abnormal values and missing values, carrying out normalization processing on the data, and scaling the data to a uniform range;
s7.2, LSTM network structure design:
The design of an input layer, namely determining the node number of the input layer, namely the parameter dimensions of a historical coupling coefficient k, an offset delta d, a temperature T and a load current I_L according to the result of data preprocessing, and taking the preprocessed data as the input of the input layer;
hidden layer design, namely adopting a 3-layer LSTM network structure;
The output layer is used for predicting the coupling coefficient k value within 10ms in the future by setting the node number of the output layer as 1, and adjusting the compensation capacitor C1/C2 through a capacitance adjustment formula according to the prediction result;
the capacitance adjustment formula is:
Wherein, Cnew is the capacitance value adjusted according to the actual demand, Cnom is the rated capacitance value marked on the capacitor, namely the design value of the capacitor under ideal conditions, and beta is the capacitance adjustment gain coefficient;
S7.3, LSTM network training:
S7.4, LSTM network prediction and compensation, namely, after training is completed, predicting a coupling coefficient k value within 10ms in the future by using the LSTM network, wherein a prediction result is used for guiding the operation and control of the system;
compensating, namely adjusting a compensating capacitor C1/C2 according to a prediction result through a capacitance adjusting formula;
S7.5, model evaluation and optimization, namely evaluating model performance, namely evaluating the predicted performance of the LSTM network by using a test data set;
An optimization model, namely adjusting and optimizing the LSTM network according to the evaluation result;
And S7.6, when the prediction accuracy reaches more than 90%, the control module is used.
Further, the method also comprises S8, wherein the magnetic circuit optimization step comprises the following steps:
S8.1, dynamically activating a segmented guide rail, namely identifying the position of a receiving end through COMSOL real-time simulation of magnetic field distribution, and activating the nearest 3-segment transmitting guide rail;
And S8.2, optimizing a magnetic line path, namely adjusting the excitation phase of the guide rail to ensure that the direction of the composite magnetic field is always aligned with the receiving end, and the fluctuation of the coupling coefficient k is less than or equal to 5 percent.
Further, S8.3, magnetic shielding self-adaptive adjustment is further included, wherein the active shielding coil is used for controlling, an auxiliary shielding coil is deployed at the edge of the receiving end, reverse current is injected according to a magnetic leakage detection signal, a leakage magnetic field is counteracted, and gradient magnetic permeability ferrite is adopted, so that the second aspect of edge magnetic flux diffusion is reduced.
Compared with the prior art, the application has at least the following beneficial effects:
1. The invention can activate the low-frequency channel by default when the system is started by a double-frequency cooperative transmission mechanism, and establish a basic energy transmission link by utilizing the strong magnetic field penetration characteristic of the low-frequency channel, and simultaneously monitor the load power requirement of a receiving end in real time, and trigger the high-frequency channel to cooperatively work if the load power requirement suddenly increases, so as to realize local enhanced transmission. The dual-frequency cooperative transmission mode ensures that the invention can maintain high-efficiency energy transmission when facing different load and offset conditions.
2. The invention can dynamically adjust the resonance capacitance and the switching frequency according to the transmission efficiency, the load impedance and the offset signal monitored in real time, so that the system always works at the optimal coupling point. In addition, the invention also adopts a neural network to predict the attenuation trend of the coupling coefficient, and adjusts the compensation capacitance value in advance so as to inhibit the efficiency fluctuation. Together, these measures enhance the stability and robustness of the system.
3. The invention adopts the composite magnetic shielding layer structure, and the technical means of dynamic activation of the segmented guide rail, optimization of the magnetic line path and the like, thereby effectively reducing the magnetic leakage intensity and the electromagnetic radiation. The wireless charging system meets the requirements of related electromagnetic radiation standards, and improves the safety and environmental protection of the wireless charging system.
4. The charging pile double-frequency coupling wireless transmission device and the efficiency compensation method are suitable for a high-dynamic electric vehicle wireless charging scene, and can keep high efficiency and low electromagnetic radiation under complex working conditions of vehicle parking position deviation, abrupt change of load power demand and the like, so that the charging pile double-frequency coupling wireless transmission device and the efficiency compensation method have wide application prospect and market value in the electric vehicle wireless charging field.
Detailed Description
The present application will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present application more apparent.
As shown in fig. 1 and fig. 2, the application provides a charging pile double-frequency coupling wireless transmission device and an efficiency compensation method, comprising a charging pile double-frequency coupling wireless transmission device, which comprises a transmitting end and a receiving end, wherein the transmitting end comprises a low-frequency channel and a high-frequency channel, the low-frequency channel drives a first coil (L1) by a low-frequency resonant circuit and has strong magnetic field penetration characteristics, the high-frequency channel drives a second coil (L2) by a high-frequency resonant circuit to realize local enhancement transmission, and a switching circuit dynamically switches a low-frequency/high-frequency excitation mode based on an electronic switch array of a Metal Oxide Semiconductor Field Effect Transistor (MOSFET);
Specifically, the low-frequency channel is driven by a low-frequency resonant circuit (20-100 kHz), and the circuit consists of a high-precision oscillator, a power amplifier and a matching network, so that stable and efficient low-frequency current can be generated.
The coil structure of the low-frequency channel is characterized in that the first coil (L1) adopts a rectangular close-wound coil structure, the outer diameter of the first coil is 1.5-2 times of that of the receiving end coil, the structure is beneficial to enhancing the focusing effect of a magnetic field, and a ferrite magnetic core is embedded to enhance the focusing of the magnetic field;
In order to reduce electromagnetic interference and heat loss, the low-frequency channel part adopts a multi-layer shielding structure and is provided with a heat dissipation system, so that long-time stable operation is ensured;
Specifically, the driving circuit of the high-frequency channel is characterized in that the second coil (L2) is driven by the high-frequency resonant circuit (6.78-13.56 MHz) and can generate high-frequency current;
The second coil (L2) adopts a grouping series wound coil structure, the turns ratio of the inner ring and the outer ring is 1:3, and the structure is beneficial to reducing high-frequency loss;
the medium filling of the high-frequency channel, namely the coil interval is filled with high dielectric constant medium (such as barium titanate ceramics), so that the efficiency of high-frequency transmission is further improved;
the function of the switching circuit is that the electronic switch array based on the MOSFET can dynamically switch the low-frequency/high-frequency excitation mode, and double-frequency cooperative transmission is realized by flexibly switching the excitation mode, so that the transmission efficiency and stability are improved;
The receiving end comprises a double-frequency pickup coil structure, wherein the receiving end comprises a tightly coupled receiving coil (L3) and a cooperative coil (L4), the L4 is tightly wound on the outer side of the L3, the cooperative coil (L4) and the transmitting end (L1/L2) form strong coupling, the coupling coefficient k of the L3 and the L4 is more than or equal to 0.4, and high-efficiency energy transmission is ensured;
The composite magnetic shielding layer is made of PC95 manganese zinc ferrite and copper foil composite shielding materials, the thickness of the shielding layer is 0.3-0.5mm, the back and the side edges of the receiving end are covered, the magnetic leakage is reduced by more than or equal to 30%, the shielding effect is ensured, the weight of the device is reduced, the magnetic leakage intensity and electromagnetic radiation are effectively reduced, and the safety of the system is improved.
The control module is used for acquiring transmission efficiency, load impedance and offset signals in real time;
the control module is internally provided with a double-frequency dynamic compensation algorithm, and adjusts the resonant capacitor (C1/C2) and the switching frequency according to the efficiency feedback, so that the system always works at an optimal coupling point;
the control module can utilize the neural network to predict the attenuation trend of the coupling coefficient k, adjust the compensation capacitance value in advance and inhibit the efficiency fluctuation;
Specifically, the control module is a core part of the dual-frequency coupling wireless transmission device and is responsible for monitoring, adjusting and optimizing a system so as to ensure the high efficiency and the stability of wireless energy transmission. The following working steps of the control module design are as follows:
Step one, data acquisition
And (3) real-time monitoring:
the control module monitors data such as offset delta d, temperature T, load current I_L and the like in real time through a high-precision sensor and an acquisition circuit, and the data are important indexes for evaluating the performance of the system and are important for optimizing and adjusting the system;
And step two, data processing, namely sending the acquired data into a microprocessor for further processing after preprocessing such as filtering, amplifying and the like, and analyzing and calculating the data by the microprocessor to obtain the current state and working parameters of the system.
Step three, working parameter input dynamic compensation algorithm
The algorithm principle is that the dynamic compensation algorithm is a closed-loop control algorithm based on system efficiency feedback, and the system always works at an optimal coupling point by adjusting the resonant capacitance (C1/C2) and the switching frequency according to the transmission efficiency monitored in real time.
The method is realized by a microprocessor, the resonance capacitance and the switching frequency are adjusted in real time according to a preset control strategy and parameters, and in the adjustment process, the algorithm considers factors such as load change, offset condition and the like of the system so as to ensure the stability and the high efficiency of the system.
The dynamic compensation algorithm can adaptively adjust system parameters, improve transmission efficiency, inhibit efficiency fluctuation caused by factors such as load change and offset, and ensure stable operation of the system.
Intelligent prediction and adjustment
The neural network prediction is that a neural network model is built in the control module and used for predicting the attenuation trend of the coupling coefficient k, and the neural network can accurately predict the change of the coupling coefficient through learning and training, so that a basis is provided for the adjustment of the system.
And the control module can adjust the compensation capacitance value in advance according to the prediction result of the neural network so as to inhibit efficiency fluctuation, and the mode of the advanced adjustment can more effectively stabilize the system performance and improve the transmission efficiency.
The control module also has an intelligent management function, and can automatically adjust control strategies and parameters according to the running state and the working environment of the system, so that the system can more flexibly cope with various changes and challenges and keep high-efficiency and stable running.
In one embodiment, as shown in fig. 2, there is also provided an efficiency compensation method, including the steps of:
S1, dual-frequency cooperative transmission, namely, initial energy transmission of a low-frequency channel (20-100 kHz), namely, default activation of the low-frequency channel when a system is started, and establishment of a basic energy transmission link by utilizing the strong magnetic field penetration characteristic of the low-frequency channel;
S2, dynamic load detection, namely monitoring the load power requirement of a receiving end in real time (through a current/voltage sensor), and triggering a high-frequency channel (6.78-13.56 MHz) to work cooperatively if the load power requirement is suddenly increased (delta P is more than 15% of rated power);
S3, starting an offset compensation mechanism, namely detecting the offset delta d of a receiving end through a Hall sensor, and gradually increasing the power ratio of a high-frequency channel when delta d is more than 5 cm to form a composite magnetic field of 'low-frequency global coverage+high-frequency local enhancement';
S4, high-frequency channel (6.78-13.56 MHz) self-adaptive power distribution, namely dynamically adjusting the ratio of high-frequency power to low-frequency power based on real-time feedback of a coupling coefficient k:
Wherein, Phigh is the allocated power of the high frequency band, Plow is the allocated power of the low frequency band, khigh is the coupling coefficient of the high frequency band, klow is the coupling coefficient of the low frequency band, RL is the return loss, rloss is the allocated loss, and alpha is the environmental correction factor (such as temperature and metal interference);
S5, pulse Width Modulation (PWM) optimization, namely adopting variable duty ratio PWM to control a high-frequency inverter, and increasing the high-frequency pulse width when shifting to compensate efficiency loss;
S6, dynamic parameter adjustment:
Real-time calculation of optimal frequency ratio based on load variationThe duty ratio of the inverter is regulated by a PID controller, wherein fhigh is high-frequency resonant frequency, flow is low-frequency resonant frequency, k is coupling coefficient, L is inductance value, RL is return loss, and Rloss is distribution loss;
PID closed loop control:
the proportional term (P) is a rapid response load mutation;
the integral term (I) eliminates steady state errors;
the differentiating term (D) suppresses the high-frequency oscillation.
PID output directly adjusts the switching frequency of the inverter, so that the system is ensured to always work at an optimal resonance point;
S7, predicting the attenuation trend of the coupling coefficient k by using a neural network, adjusting a compensation capacitance value in advance, and inhibiting efficiency fluctuation, wherein the neural network predicts the attenuation trend of the coupling coefficient k and comprises the following steps:
S7.1, collecting and preprocessing data of parameters such as a historical coupling coefficient k, an offset delta d, a temperature T, a load current I_L and the like, wherein the data cover the running states of the system under different working conditions and different environmental conditions so as to ensure that an LSTM network can learn the overall system characteristics;
the data preprocessing comprises the steps of cleaning the collected data, removing abnormal values, missing values and the like, normalizing the data, and scaling the data to a uniform range so as to improve the training efficiency and the prediction accuracy of an LSTM network;
s7.2, designing an LSTM network structure, wherein the design of an input layer comprises the steps of determining the node number of the input layer, namely the dimensionality of parameters such as a historical coupling coefficient k, an offset delta d, a temperature T, a load current I_L and the like according to the result of data preprocessing, and taking the preprocessed data as the input of the input layer;
The hidden layer design is that a 3-layer LSTM network structure is adopted, the number of LSTM units in each layer can be adjusted according to the complexity and the prediction requirement of the system, the LSTM network can process time sequence data, and the cell state is controlled and updated through a forgetting gate, an input gate and an output gate in the LSTM network, so that the long-term prediction of the system state is realized;
the output layer is used for predicting the coupling coefficient k value within 10ms in the future by setting the node number of the output layer as 1, and adjusting the compensation capacitance C1/C2 through a capacitance adjustment formula according to the prediction result;
the capacitance adjustment formula is:
wherein, Cnew is the capacitance value adjusted according to the actual demand, Cnom is the rated capacitance value marked on the capacitor, i.e. the design value of the capacitor under ideal conditions, and beta is the capacitance adjustment gain coefficient (experimental calibration);
S7.3, LSTM network training:
s7.3.1 selecting a training algorithm and an optimizer:
selecting a back propagation algorithm (BPTT), selecting an Adam optimizer for updating weights and biases of the LSTM network;
S7.3.2 setting training parameters, namely determining training parameters such as training round number (epochs), batch size (batch size) and the like, wherein the parameters influence the training efficiency and the prediction performance of the LSTM network.
S7.3.3, training, namely inputting the preprocessed data into an LSTM network for training, monitoring the change of a loss function in the training process, and adjusting training parameters to optimize the network performance;
s7.4, LSTM network prediction and compensation:
And predicting the coupling coefficient k value within 10ms in the future by using an LSTM network after training is finished, wherein the prediction result can be used as a priori information of the system state and used for guiding the operation and control of the system.
And (3) implementing compensation, namely adjusting the compensation capacitor C1/C2 through a capacitance adjustment formula according to the prediction result, wherein the adjustment of the compensation capacitor can be performed in real time so as to compensate the efficiency loss caused by the change of the system state.
S7.5 model evaluation and optimization
Evaluating model performance using the test data set to evaluate the predicted performance of the LSTM network, the evaluation index may include prediction accuracy, mean Square Error (MSE), etc.
And (3) optimizing the model, namely adjusting and optimizing the structure, training parameters and the like of the LSTM network according to the evaluation result, and improving the prediction performance and compensation effect of the LSTM network by continuously iterating and optimizing.
And S7.6, when the prediction accuracy reaches more than 90%, the control module is used.
S8, magnetic circuit optimization and anti-offset control:
S8.1, dynamically activating a segmented guide rail, namely performing finite element simulation feedback, namely identifying the position of a receiving end through COMSOL real-time simulation of magnetic field distribution, and activating the nearest 3-segment transmitting guide rail (each segment length=1.2 times of a coil of the receiving end);
And S8.2, optimizing a magnetic line path, namely adjusting the excitation phase of the guide rail to ensure that the direction of the composite magnetic field is always aligned with the receiving end, and the fluctuation of the coupling coefficient k is less than or equal to 5 percent.
S8.3, magnetic shielding self-adaptive adjustment, namely, active shielding coil control, namely, arranging an auxiliary shielding coil at the edge of a receiving end, injecting reverse current according to a magnetic leakage detection signal (B < 27 mu T), and counteracting a leakage magnetic field;
material optimization, namely gradient magnetic permeability ferrite (mu_r is gradually changed from 1000 to 3000) is adopted, and edge magnetic flux diffusion is reduced.
Experimental verification and effect
| Index (I) | Traditional single frequency system | The scheme of the invention | Amplitude of lift |
| Maximum offset tolerance | 10 cm | 20 cm | 100% |
| Efficiency fluctuation (Deltaeta) | ±15% | ±3% | 80% |
| Leakage intensity (@ 1 m) | 35 μT | 22 μT | 37% |
| Dynamic response time | 50 ms | 5 ms | 90% |
The conclusion is that the system can still keep high efficiency (more than 90%) and low electromagnetic radiation under complex working conditions through double-frequency cooperative transmission, intelligent parameter adjustment and magnetic circuit optimization, and the system is suitable for a high-dynamic electric automobile wireless charging scene.
One specific application example is also given below:
the size of the transmitting end L1 is 60 cm multiplied by 40 cm (low frequency), the size of the receiving end L2 is 30 cm multiplied by 30 cm (high frequency), and the receiving end L3/L4 adopts 20 cm multiplied by 20 cm double-layer concave-convex magnetic cores;
the control module takes STM32H7 as a core, the sampling frequency is 1 kHz, and the C1/C2 range is dynamically adjusted to 10-100 nF;
experiments show that when the transverse offset is 15 cm, the transmission efficiency is improved to 89% from 75% of the traditional single-frequency system, and the magnetic leakage intensity is less than or equal to 25 mu T (meeting the GB/T38775.4 standard).
The technical features of the above embodiments may be arbitrarily combined, and all possible combinations of the technical features in the above embodiments are not described for brevity of description, however, as long as there is no contradiction between the combinations of the technical features, they should be considered as the scope of the description.