Load and mutual inductance synchronous identification method for multi-frequency and multi-load wireless power transmission system

Abstract

The application discloses a method for load and mutual inductance synchronization identification of a multi-frequency and multi-load wireless power transmission system. First, input impedance equation sets under different frequencies are established under system steady state to obtain a relationship expression of mutual inductance and load. Meanwhile, a sliding window DFT is used to extract a frequency component of a primary side multi-frequency superimposed current according to system working frequencies. Then, a particle swarm algorithm is designed by taking an error between an actual output voltage value and a theoretical value as a fitness function, a parameter identification problem is converted into an algorithm optimization problem, and an optimal solution search is performed on the parameter to be identified to replace a traditional calculation method, so that errors generated by the traditional equation are avoided. The application only needs to analyze a model of system input impedance, has a simple circuit structure, does not need an additional control circuit, and only needs to sample an input bus voltage, a primary side current and an output voltage, so that communication data volume can be effectively reduced, circuit complexity is reduced, system volume is reduced, and a working range of a wireless charging system is expanded.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of wireless charging, and particularly relates to a method for load and mutual inductance identification of a multi-frequency and multi-load wireless power system. BACKGROUND

[0002] In recent years, wireless power transfer (WPT) is gradually applied in medium and low power electronic devices, and magnetic coupling resonance wireless power transfer (MCR-WPT) has become one of the main development directions of wireless power transmission due to the advantages of long transmission distance, high transmission power and high safety. In the development process of wireless power transmission, multi-load systems have been widely concerned due to their industrial significance.

[0003] Whether single load or multi-load, the most likely change in the use of a wireless charging system is the relative position of the resonator and the impedance of the connected device. The transmission performance of the system will be reduced or even out of control due to resonator offset, equivalent impedance change of the load, etc. To ensure that the system can achieve high transmission performance, real-time correction of the system model is needed, and a method for accurately identifying mutual inductance and load is needed to improve the accuracy of the model.

[0004] At present, there is a method for identifying the load and mutual inductance of a multi-load wireless power transmission system based on time division multiplexing. This method converts the multi-load system into single-load systems working at different times, and realizes the time sequence division of the multi-load. The identification process needs to collect the bus voltage and current of the system and the DC output voltage. This method essentially belongs to parameter identification of a single-load wireless charging system and cannot be applied to the application scenario of a multi-load wireless charging system. At the same time, additional circuit adjustment of the bus voltage is needed to prevent overshoot during the identification process, which is difficult to control and implement.

[0005] Therefore, how to provide a multi-frequency and multi-load wireless power transmission system load and mutual inductance synchronous identification method to realize real-time identification of the size of multiple parameters of a multi-load wireless charging system and further expand the working range of the wireless charging system has become a problem to be solved. SUMMARY

[0006] The purpose of the present application is to provide a multi-frequency and multi-load wireless power transmission system load and mutual inductance synchronous identification method, which can realize real-time identification of the size of the load and mutual inductance of a multi-load wireless charging system and further expand the working range of the wireless charging system.

[0007] The multi-frequency and multi-load wireless power transmission system load and mutual inductance synchronous identification method of the present application is applied to a multi-load magnetic coupling resonance system and includes the following steps:

[0008] Step 10, obtain a multi-frequency driving signal by superimposing different frequency modulation waves and comparing the sine pulse width modulation mode of the carrier wave, inject energy of multiple different frequencies to the multi-load wireless charging system based on the multi-frequency resonance compensation network, and transmit the power to multiple loads through the multi-frequency resonance compensation network of the transmitting end;

[0009] Step 20, obtain the relationship expression between the load and the mutual inductance by modeling the system input impedance at different frequencies;

[0010] Step 30, use the sliding window DFT to extract the primary side multi-frequency superimposed current according to the system operating frequency component, sample the system input bus voltage, primary side current and DC output voltage in real time, calculate the identified model output voltage calculation value through the data obtained by sampling and the charging parameters of the system, and then take the error between the actual system output voltage value and the identified model output voltage calculation value as the objective function for evaluating the fitness of the particle;

[0011] Step 40, optimize the target function by using the particle swarm algorithm, take the global optimal solution searched at the end of the operation of the particle swarm algorithm as the load value, and then determine the mutual inductance value according to the relationship expression between the mutual inductance and the load, so as to realize the synchronous identification of the load and the mutual inductance of the multi-load wireless charging system in real time.

[0012] The step 20 comprises:

[0013] Step 21, Fourier series expansion is performed on the output voltage after inversion, and the fundamental wave wherein i = 1, 2,..., n, is the different frequency inverter output voltage component, a i is the modulation ratio, E dc is the DC bus voltage; i represents any modulation ratio, any operating frequency and the receiving end loop or load, and n represents the number of parallel LC networks;

[0014] Step 22, the equivalent impedance of the transmitting end multi-resonant frequency compensation network is:

[0015]

[0016] wherein ω is the system operating angular frequency, L pn is the primary side compensation inductance, C pn is the primary side compensation capacitance;

[0017] The relationship expression between the load and the mutual inductance, taking a double-frequency double-load MCR-WPT system as an example, the input impedance equation group of the system working at different frequencies is:

[0018]

[0019] where R1=R eq1 +R s1 , R2=R eq2 +R s2 , μ=ω1L s2 -1 / (ω1C s2 ), β=ω2L s1 -1 / (ω2C s1 ), and are input impedance at different frequencies, L s1 and L s2 are secondary side compensation inductance, C s1 and C s2 are secondary side compensation capacitance, R p is primary side coil resistance, R s1 and R s2 are secondary side coil resistance, R L1 and R L2 are secondary side load, M ps1 and M ps2 are coil mutual inductance, ω1 and ω2 are system working angular frequency;

[0020] According to the input impedance equation set of the system working at different frequencies, the relationship expression between the load and the mutual inductance can be obtained:

[0021]

[0022] where,

[0023] The step 30 comprises:

[0024] Step 31, the fundamental wave current expression extracted by the sliding window DFT is:

[0025]

[0026] Wherein, I p (n) is the signal value at n time, N is the sampling point number of a period, A1 and B1 are the real part and imaginary part coefficients of the current expression, that is:

[0027]

[0028] Wherein, N cur is the current sampling point, N cur -N+1 is the initial sampling point;

[0029] Step 32, the voltage calculation value output by the identification model is:

[0030]

[0031] Wherein, Cfi P is a rectifier-side post-stage filter capacitor i P is a system output power, T is a running time, V i Vi is an i-th load output voltage; R Li represents an i-th load; k represents a k-th moment of T;

[0032] Step 33, taking an actual system output voltage value V i * and an error of an identified model output voltage calculation value V i as a target function to evaluate a particle fitness:

[0033]

[0034] The step 40 comprises:

[0035] Step 41, setting an iteration threshold value, and taking a mutual inductance value between a transmitting end and a receiving end as a particle input particle swarm algorithm;

[0036] Step 42, initializing a speed and a position of each particle;

[0037] Step 43, calculating an inertia weight factor of each particle;

[0038] Step 44, updating a speed and a position of each particle based on the inertia weight factor;

[0039] Step 45, calculating a fitness of each particle based on the target function, and determining an individual optimal value and a global optimal value of the particle based on the fitness;

[0040] Step 46, outputting the global optimal value based on the iteration threshold value and the fitness, and taking the global optimal solution as a load value;

[0041] Step 47, simultaneously obtaining sizes of the mutual inductance and the load according to a relationship expression between the load and the mutual inductance.

[0042] The step 45 comprises:

[0043] Step 451, calculating a fitness (n) of each particle in the n-th iteration based on the target function;

[0044] Step 452, judging whether the fitness (n) is less than the fitness (n-1), if yes, setting the individual optimal value gbest(n)=fitness(n), and entering step 453; if no, entering step 453;

[0045] Step 453, judging whether the fitness(n) is less than the global optimal value zbest(n), if yes, setting the global optimal value zbest(n)=fitness(n), and entering step 454; if no, entering step 454;

[0046] Step 454, after the iteration number n is added by 1, entering step 46.

[0047] The technical scheme of the present application firstly establishes the system input impedance equation set under different frequencies under the system steady state, and obtains the mutual inductance relationship expression about the load; at the same time, the original side multi-frequency superimposed current is divided and extracted by using the sliding window DFT; then the error between the actual value and the theoretical value of the output voltage is used as the fitness function to design the particle swarm algorithm, the parameter identification problem is converted into the algorithm optimization problem, and the optimal solution search is performed on the to-be-identified parameters to replace the traditional calculation method, thereby avoiding the error generated by the traditional equation. The present application only needs to analyze the model of the system input impedance, the circuit structure is simple, no additional control circuit is needed, and only the input bus voltage, the primary side current and the output voltage need to be sampled, which can effectively reduce the communication data amount, reduce the circuit complexity and reduce the system volume, and expand the working range of the wireless charging system. BRIEF DESCRIPTION OF DRAWINGS

[0048] Fig. 1(a) is a main circuit diagram of the multi-load wireless charging system of the present application;

[0049] Fig. 1(b) is an original side circuit diagram based on the multi-frequency compensation network of the multi-load wireless charging system of the present application;

[0050] Figure 2 Fig. 2 is a total framework diagram of the multi-load wireless charging system in the present application;

[0051] Figure 3 Fig. 3 is a mutual inductance and load synchronous identification flow chart in the present application;

[0052] Figure 4 Fig. 4 is a sliding window DFT current detection flow chart in the present application.

[0053] The present application will be further described in detail below in combination with the drawings and specific embodiments. DETAILED DESCRIPTION

[0054] The embodiment of the present application provides a multi-frequency multi-load wireless power transmission system load and mutual inductance synchronous identification method, realizes real-time identification of the load and the mutual inductance size of the multi-load wireless charging system, and further expands the working range of the wireless charging system.

[0055] The multi-load wireless charging system based on a multi-frequency resonance compensation network in the embodiment of the application is a multi-frequency energy wireless power transmission system in which a multi-frequency resonance compensation network is designed in a transmitting end circuit to realize multi-frequency energy wireless power transmission. The general idea of the technical solution is as follows: first, input impedance equation sets at different frequencies are established under system steady state, and a mutual inductance and load relationship expression is obtained by solving; at the same time, the original side multi-frequency superimposed current is divided and extracted by using a sliding window DFT; then, the actual value and the theoretical value error of the output voltage are used as the fitness function to design a particle swarm algorithm, the parameter identification problem is converted into an algorithm optimization problem, and the optimal solution of the to-be-identified parameters is searched instead of the traditional calculation method, thereby avoiding the error generated by the traditional equation. The method only needs to analyze the system input impedance model, and the circuit structure is simple without additional control circuit. The method only needs to sample the input bus voltage, the primary side current and the output voltage, thereby reducing the communication data volume, reducing the circuit complexity and reducing the system volume, and expanding the working range of the wireless charging system.

[0056] The multi-load wireless charging system based on a multi-frequency resonance compensation network in the embodiment of the application, as shown in FIGS. 1(a) and (b), includes a voltage-stabilized DC power supply, a full-bridge inverter (S1-S4), a resonator of the transmitting end, and a rectification and filtering module and a load of the receiving end. The voltage-stabilized DC power supply converts and outputs high-frequency alternating current through the full-bridge inverter, and then transmits energy to multiple receiving ends through the resonator through a magnetic field, and finally transmits the energy to the load through the rectification and filtering module. The full-bridge inverter is composed of GaN-MOSFET. The resonator is composed of a symmetric circular coil wound by a Litz wire, and includes a transmitting end resonant circuit and a receiving end resonant circuit. The transmitting end of the multi-load wireless charging system uses a full-bridge inverter to convert a DC power supply to generate a high-frequency alternating voltage. The rectification and filtering module of the receiving end uses an uncontrollable bridge rectifier to simplify the control difficulty. The resonator is composed of n-1 parallel LC networks and a series inductor and capacitor.

[0057] (1) The transmitting end resonant circuit adopts a multi-frequency resonance compensation network topology design

[0058] The transmitting end resonant circuit of the application adopts a multi-frequency resonance compensation network that can independently control the input phase at n resonance frequencies, as shown in FIG. 1(b). The multi-frequency resonance compensation network is composed of one capacitor and n-1 parallel LC circuits. When the multi-load wireless charging system works in a resonant state, the phase angle of the system input impedance at the corresponding resonance frequency is zero, and the multi-frequency reactive component can be completely eliminated.

[0059] (2) Sliding window DFT design

[0060] As Figure 4As shown, the present application detects the current signal of each frequency component in the primary side circuit through the sliding window DFT current detection method, only needs to perform iterative calculation on the current sampling point, and does not need to perform iterative calculation on the sampling points in the whole period, thereby reducing the calculation amount and calculation time.

[0061] (3) Recognition module design

[0062] In order to realize the multi-parameter identification of the multi-frequency multi-load wireless charging system, the present application proposes a representative expression of the mutual inductance and the load under different frequencies of the system input impedance, takes the error between the actual value and the theoretical value of the output voltage as the fitness function of the particle swarm algorithm, takes the global optimal solution searched at the end of the operation of the particle swarm algorithm as the load value, and then determines the mutual inductance value according to the expression of the mutual inductance and the load. In this process, the particle swarm algorithm only needs to identify one parameter of the load, so the number of identified parameters is small, and the identification time is short. The identification process is as shown in Figure 3 .

[0063] As shown in Figures 2 to 4 , the present application is a method for synchronous identification of the load and the mutual inductance of a multi-frequency multi-load wireless power transmission system, applied to a multi-load magnetic coupling resonance system, comprising the following steps:

[0064] Step 10: Obtain a multi-frequency driving signal through a sine pulse width modulation mode of superimposing different frequency modulation waves and a carrier wave comparison, inject energy of multiple different frequencies into a multi-load wireless charging system based on a multi-frequency resonance compensation network, and transmit the power to multiple loads through the multi-frequency resonance compensation network in the transmitting end;

[0065] Step 20: Model the system input impedance under different frequencies to obtain an expression of the relationship between the load and the mutual inductance;

[0066] Step 30: Extract the primary side multi-frequency superimposed current according to the system working frequency component through sliding window DFT, sample the system input bus voltage, the primary side current and the direct current output voltage in real time, calculate the identification model output voltage value through the data obtained by sampling and the charging parameters of the system, and then take the error between the actual system output voltage value and the identification model output voltage calculation value as the target function for evaluating the fitness of the particle;

[0067] Step 40: Optimize the target function by using the particle swarm algorithm, take the global optimal solution searched at the end of the operation of the particle swarm algorithm as the load value, and then determine the mutual inductance value according to the expression of the mutual inductance and the load, so as to realize the synchronous identification of the load and the mutual inductance of the multi-load wireless charging system in real time.

[0068] The step 20 comprises:

[0069] Step 21: Fourier series expansion is performed on the output voltage after inversion, and the fundamental wave is taken where i = 1, 2, …, n, are different frequency inverse output voltage components, a i is the modulation ratio, E dc is the DC bus voltage; i represents any modulation ratio, any operating frequency and receiving end circuit or load, and n represents the number of parallel LC networks;

[0070] The equivalent impedance of the transmitting end multi-resonant frequency compensation network is:

[0071]

[0072] where ω is the system operating angular frequency, L pn is the primary side compensation inductance, C pn is the primary side compensation capacitance;

[0073] The relationship expression between the load and mutual inductance is expressed, taking a double-frequency double-load MCR-WPT system as an example, and assuming that the input impedance equation set of the system operating at different frequencies is:

[0074]

[0075] In the formula, R1 = R eq1 + R s1 , R2 = R eq2 + R s2 , μ = ω1L s2 - 1 / (ω1C s2 ), and β = ω2L s1 - 1 / (ω2C s1 ). and are input impedances at different frequencies, L s1 and L s2 are secondary side compensation inductances, C s1 and C s2 are secondary side compensation capacitances, R p is the primary side coil resistance, R s1 and R s2 are secondary side coil resistances, R L1 and R L2 are secondary side loads, M ps1 and M ps2 are coil mutual inductances, and ω1 and ω2 are system operating angular frequencies;

[0076] According to the input impedance equation set of the system operating at different frequencies, the relationship expression between the load and mutual inductance can be obtained:

[0077]

[0078] In the formula,

[0079] Step 30 includes:

[0080] Step 31: The expression for the fundamental current extracted by the sliding window DFT is:

[0081]

[0082] Among them, I p (n) represents the signal value at time n, N is the number of sampling points in one period, and A1 and B1 are the real and imaginary coefficients of the current expression, respectively:

[0083]

[0084] In the formula, N cur For the current sampling point, N cur -N+1 represents the initial sampling points;

[0085] Step 32, the calculated output voltage value of the identification model is:

[0086]

[0087] Among them, C fi For the rectifier-side post-stage filter capacitor, P i V represents the system output power, T represents the running time, and V represents the operating time. i R is the output voltage of the i-th load; Li Represents the i-th load; k represents the k-th time in time T;

[0088] Step 33: Convert the actual system output voltage value V i * The calculated output voltage V of the identification model i The error is used as the objective function to evaluate the particle's fitness:

[0089]

[0090] Step 40 includes:

[0091] Step 41: Set the iteration threshold and use the mutual inductance value between the transmitter and receiver as the particle input to the particle swarm algorithm;

[0092] Step 42: Initialize the velocity and position of each particle;

[0093] Step 43: Calculate the inertia weighting factor for each particle;

[0094] Step 44: Update the velocity and position of each particle based on the inertia weighting factor;

[0095] Step 45, calculating the fitness of each particle based on the objective function, and determining the individual optimal value and the global optimal value of the particle based on the fitness;

[0096] Step 46, outputting the global optimal value based on the iteration threshold value and the fitness, and taking the global optimal solution as the load value;

[0097] Step 47, obtaining the size of the mutual inductance and the load according to the relationship expression between the load and the mutual inductance.

[0098] The step 45 comprises:

[0099] Step 451, calculating the fitness (n) of each particle in the nth iteration based on the objective function;

[0100] Step 452, judging whether the fitness (n) is less than the fitness (n-1), if yes, taking the individual optimal value gbest (n) = fitness (n), and entering step 453; if not, entering step 453;

[0101] Step 453, judging whether the fitness (n) is less than the global optimal value zbest (n), if yes, taking the global optimal value zbest (n) = fitness (n), and entering step 454; if not, entering step 454;

[0102] Step 454, after the iteration number n is added by 1, entering step 46.

[0103] The technical focus of the application is:

[0104] The method for realizing the directional transmission of the power of the multi-load wireless power transmission system by increasing the multi-resonant frequency compensation network on the primary side, without additional control circuit, and the cross-coupling influence between the receiving ends is small and can be ignored;

[0105] In view of the existence of the multi-frequency superimposed current in the primary side circuit, the current signal of the corresponding working frequency needs to be sampled in the identification process, and therefore the sliding window DFT current detection method is adopted to realize the extraction of the fundamental component of the primary side current signal corresponding to the frequency;

[0106] The relationship expression of the mutual inductance and the load is calculated by establishing the input impedance of the system working at different frequencies, the particle swarm algorithm is introduced, and the difference between the sampled output voltage and the predicted output voltage is used to create the fitness function, the parameter identification problem is converted into the algorithm optimization problem, the optimal solution search of the identification parameter is used to replace the traditional calculation method, and the error generated by the traditional equation is avoided;

[0107] By sampling the power bus voltage, primary side current and DC output voltage of the wireless charging system in real time, the high-frequency large voltage across the devices in the coupling mechanism does not need to be directly measured, which is safer, and the algorithm has low complexity, short operation time, and small error.

[0108] After the technical solution of the application is adopted, when mutual inductance is disturbed and deviates from the set value, the size between the mutual inductance and the load can be effectively identified and the value in the prediction model can be corrected; through the cooperation of the model predictive control algorithm (MPC algorithm) and the particle swarm optimization algorithm (PSO algorithm), the dynamic stability and fast response capability of the wireless charging system can be ensured; that is, by combining the PSO algorithm with the MPC algorithm, the application can be used in both dynamic and static conditions, making the wireless charging more reliable; not only can the receiving end imaginary part estimation in the offline state be realized, but also the receiving end imaginary part estimation of the dynamic wireless charging system can be realized, greatly improving the practicability of the application.

[0109] Although the specific embodiments of the application are described above, those skilled in the art should understand that the specific examples described are only illustrative, and are not intended to limit the scope of the application, and equivalent modifications and changes made by those skilled in the art in accordance with the spirit of the application should be covered by the scope of the claims of the application.

Claims

1.A method for load and mutual inductance synchronization identification of a multi-frequency multi-load wireless power transmission system, applied to a multi-load magnetic coupling resonance system, characterized in that It comprises the following steps: Step 10, a multi-frequency driving signal is obtained by superimposing different frequency modulation waves and comparing the sine pulse width modulation mode of the carrier wave, energy of different frequencies is injected into the multi-load wireless charging system based on the multi-frequency resonance compensation network, and the transmitting end transmits power to the multiple loads through the multi-frequency resonance compensation network; Step 20, by modeling the system input impedance at different frequencies, the relationship expression between the load and the mutual inductance is obtained; Step 30, the primary multi-frequency superimposed current is divided into frequency components according to the system operating frequency, the system input bus voltage, the primary current and the DC output voltage are sampled in real time, the identified model output voltage calculation value is calculated by using the data obtained by sampling and the charging parameters of the system, and then the error between the actual system output voltage value and the identified model output voltage calculation value is used as the objective function to evaluate the fitness of the particles; Step 40, the particle swarm algorithm is used to optimize the target function, and the global optimal solution searched at the end of the operation of the particle swarm algorithm is used as the load value, and then the mutual inductance value is determined according to the relationship expression between the mutual inductance and the load, so that the load and the mutual inductance of the multi-load wireless charging system are identified in real time; The step 20 comprises: Step 21, Fourier series expansion of the output voltage after inversion, taking its fundamental wave wherein , is the output voltage component of different frequency inversion, is the modulation ratio, is the DC bus voltage; i represents any modulation ratio, any operating frequency and the receiving end circuit or load, and n represents the number of parallel LC networks; Step 22, the equivalent impedance of the transmitting end multi-resonant frequency compensation network is: wherein, is a system operating frequency, is a primary side compensation inductance, is a primary side compensation capacitance; The relationship expression between the load and the mutual inductance, under the double-frequency double-load MCR-WPT system, the input impedance equation group of the system working at different frequencies is: In the formulae, , and is the input impedance at different frequencies, and is the secondary side compensation inductance, and is the secondary side compensation capacitance, is the primary side coil resistance, and is the secondary side coil resistance, and is the secondary side load, and is the coil mutual inductance, and is the system operating angular frequency at two frequencies; According to the input impedance equation group of the system working at different frequencies, the relationship expression between the load and the mutual inductance is obtained: In the formulae, ; The step 30 comprises: Step 31, the fundamental current expression extracted by the sliding window DFT is: wherein, is the signal value at time m, N is the number of sampling points in one period, and A1 and B1 are the real and imaginary coefficients of the current expression, i.e.: wherein is the current sample point, is the initial sample point; Step 32, the identified model output voltage calculation value is: wherein, is a rectifier-side post-stage filter capacitor, is a system output power, T is a running time, is a jth load output voltage; represents a jth load; k represents a kth time of T; Step 33, calculating the actual system output voltage value the error between the recognized model output voltage calculated value as a target function for evaluating the fitness of the particles: 。 2. The method of claim 1, wherein, It comprises the following steps: The step 40 comprises: Step 41, set the iteration threshold value, and input the mutual inductance value between the transmitting end and the receiving end into the particle swarm algorithm as a particle; Step 42, initialize the speed and position of each particle; Step 43, calculate the inertia weight factor of each particle; Step 44, update the speed and position of each particle based on the inertia weight factor; Step 45, calculate the fitness of each particle based on the objective function, and determine the individual optimal value and the global optimal value of the particle based on the fitness; Step 46, output the global optimal value based on the iteration threshold value and the fitness, and take the global optimal solution as the load value; Step 47, the mutual inductance and the load are obtained according to the relationship expression between the load and the mutual inductance; The step 45 comprises: Step 451, calculating the fitness of each particle in the n-th iteration based on the target function ; Step 452, judging whether it is less than , if yes, setting the individual optimal value , and entering step 453; if not, entering step 453; Step 453, judging whether it is less than the global optimal value , if yes, setting the global optimal value , and entering step 454; if no, entering step 454; Step 454, after the iteration number n is increased by 1, go to step 46.

Images