Multi-frequency multi-load wpt system based on parameter identification and control method thereof

Abstract

The application relates to the technical field of wireless power transmission, and particularly discloses a multi-frequency multi-load WPT system based on parameter identification and a control method thereof. The size of load and mutual inductance parameters can be calculated by directly and accurately measuring the primary side current, so that the identification of the load and mutual inductance parameters of the multi-frequency multi-load WPT system can be realized without affecting the normal operation of the system, that is, the parameter identification and power transmission of the multi-frequency multi-load system can be simultaneously realized. The frequency and power adjustment are convenient, can be adjusted according to requirements, and meet the needs of the load. The load and mutual inductance parameters are identified based on the NSGA II algorithm and the TrustRegion algorithm, the identification accuracy is high, and no additional controller is needed. The problems that the traditional parameter identification method can only identify a single load or mutual inductance parameter, the result is inaccurate, the circuit is complex, and detection is difficult are solved.

Description

Technical Field

[0001] This invention relates to the field of wireless power transfer (WPT) technology, and in particular to a multi-frequency, multi-load WPT system and its control method based on parameter identification. Background Technology

[0002] Traditional power transmission methods typically use cables and other connecting devices to allow electrical equipment to draw power from the power grid. These connections are exposed and subject to friction, which can cause damage and electrical sparks. This not only reduces the lifespan of electrical equipment but also threatens the safety of the power supply, especially in harsh environments such as mines and oil fields, potentially leading to major safety accidents.

[0003] To address the aforementioned problems, wireless power transfer (WPT) technology has emerged. WPT technology allows for a direct connection between the power grid and electrical equipment without the need for cables or other connecting devices, significantly overcoming the shortcomings of traditional power supply methods. Among these, WPT technology is currently the most extensively researched wireless power transfer method. However, as research progresses, researchers have increasingly focused on the adaptability of system parameters. The mutual inductance and load parameters of WPT systems exhibit variable and unknown characteristics. Changes in these parameters can cause the system's operating frequency to drift, deviating from its resonant frequency and reducing system stability, thus posing challenges to WPT system research.

[0004] To address this challenge, researchers have conducted extensive studies. Currently, the methods employed can be broadly categorized into five types: impedance matching methods, energy conservation methods, intelligent algorithm identification methods, identification methods using external auxiliary control devices, and parameter identification methods based on fundamental and harmonic frequencies. However, these methods all suffer from several insurmountable drawbacks: the first and second types, while achieving load parameter identification, are based on single-frequency, single-load MCR-WPT systems; the third type is often computationally complex; the fourth type requires external control devices, increasing control complexity, system size, and cost; and while the fifth type can achieve some degree of accurate load parameter identification, its fixed harmonic frequencies make it difficult to simultaneously identify more parameters. Furthermore, a common flaw in existing WPT system mutual inductance and load parameter identification algorithms is the inability to perform power transfer simultaneously during parameter identification; that is, parameter identification and power transfer cannot be carried out concurrently, making dynamic identification of mutual inductance and load parameters impossible. Summary of the Invention

[0005] This invention provides a multi-frequency, multi-load WPT system and its control method based on parameter identification. The technical problem it solves is: how to dynamically identify mutual inductance and load parameters under normal power supply mode, and how to control the system based on the identification results.

[0006] To solve the above technical problems, the present invention provides a multi-frequency multi-load WPT system based on parameter identification, including a DC power supply 1, a high-frequency inverter 2, a primary-side energy transmission module 3, a secondary-side energy receiving module 4, a system load 5, and a primary-side signal control and processing module 6.

[0007] The secondary side energy receiving module 4 includes n≥2 natural resonant frequencies f1, f2, ... f1, f2, ... n The power receiving circuit includes a secondary magnetic energy pickup coil and a secondary compensation capacitor connected in series.

[0008] The primary-side signal control and processing module 6 includes a current detection circuit, a DSP module, and a composite modulation PWM control circuit connected in sequence.

[0009] The composite modulation PWM control circuit is used to output n composite modulation waves of different frequencies according to n different inherent resonant frequencies of the power receiving circuit to drive the high-frequency inverter 2 to work.

[0010] The current detection circuit is used to acquire the transmitter-side current; the DSP module is used to perform FFT analysis on the transmitter-side current acquired by the current detection circuit to obtain the effective value I of the transmitter-side current at different frequencies. p1mea ,I p2mea …I pnmea and the phase difference θ1, θ2…θ with respect to voltage n tangent value tanθ 1mea ,tanθ 2mea …tanθ nmea , and according to I p1mea ,I p2mea …I pnmea and tanθ 1mea ,tanθ 2mea …tanθ nmea Calculate the identification parameter M ps1 M ps2 …M psn and R L1 ,R L2 …R Ln M psi This indicates that the primary side energy emission module 3 and the i-th secondary side magnetic energy pickup coil L si Mutual intuition between them, R Li This refers to the load in load 5 of the system that is connected to the i-th secondary magnetic energy pickup coil, where i = 1, 2, ..., n;

[0011] The DSP module is also used to identify parameter M ps1 M ps2 …M psn and R L1 ,RL2 …R Ln The composite modulation PWM control circuit outputs a new composite modulation wave to change the operating state of the high-frequency inverter 2 so that it adapts to the current load requirements.

[0012] Preferably, the DSP module is based on I p1mea ,I p2mea …I pnmea and tanθ 1mea ,tanθ 2mea …tanθ nmea Calculate the identification parameter M ps1 M ps2 …M psn and R L1 ,R L2 …R Ln The specific process is as follows:

[0013] Identification parameter M ps1 M ps2 …M psn and R L1 ,R L2 …R Ln As the optimization variable for the multi-objective optimization problem, the effective value of the primary current I... p1 ,I p2 …I pn The measured value I p1mea ,I p2mea …I pnmea Compared with the theoretically calculated value I p1 (ψ),I p2 (ψ)…I pn By subtracting (ψ) from each other, we construct the first objective function F(ψ) and the first optimization problem:

[0014] P1:

[0015] stM psiL <M psi <M psiH ,R LiL <R Li <R LiH i = 1, 2, ..., n

[0016] Among them, M psiL M psiH M respectively psi The upper and lower limits of R LiL R LiH R respectively Li The upper and lower limits;

[0017] Solve the first optimization problem to obtain the identification parameters M.ps1 M ps2 …M psn and R L1 ,R L2 …R Ln The initial solution.

[0018] Preferably, the DSP module is further used to perform FFT analysis on the transmitter-side current acquired by the current detection circuit to obtain the phase differences θ1, θ2…θ between the transmitter-side current and voltage at different frequencies. n tangent value tanθ 1mea ,tanθ 2mea …tanθ nmea It is also used according to I p1mea ,I p2mea …I pnmea tanθ 1mea ,tanθ 2mea …tanθ nmea and identification parameter M ps1 M ps2 …M psn and R L1 ,R L2 …R Ln The initial solution is used to calculate the identification parameters M. ps1 M ps2 …M psn and R L1 ,R L2 …R Ln The final solution.

[0019] Preferably, the DSP module is based on I p1mea ,I p2mea …I pnmea tanθ 1mea ,tanθ 2mea …tanθ nmea and identification parameter M ps1 M ps2 …M psn and R L1 ,R L2 …R Ln The initial solution is used to calculate the identification parameters M. ps1 M ps2 …M psn and R L1 ,R L2 …R Ln The specific process of the final solution is as follows:

[0020] Identification parameter M ps1 M ps2 …M psn and R L1 ,RL2 …R Ln Using the initial solution as the initial value, the effective value of the primary current I p1 ,I p2 …I pn The measured value I p1mea ,I p2mea …I pnmea Compared with the theoretically calculated value I p1 (ψ),I p2 (ψ)…I pn (ψ) are calculated separately, and the tangent of the phase difference between the primary voltage and the current is measured as tanθ. 1mea ,tanθ 2mea …tanθ nmea By subtracting the theoretically calculated values ​​tanθ1(ψ), tanθ1(ψ)…tanθ1(ψ) respectively, a second objective function λ(ψ) is constructed, and a second optimization problem is established:

[0021] P2:

[0022] Expanding the second objective function λ(ψ) using a second-order Taylor expansion yields a quadratic approximation model:

[0023]

[0024] st||d||≤Δ k

[0025] Where d=ψ-ψ k Indicates the trial step size, g k The second objective function λ(ψ) at the current iteration point ψ k gradient at point B k Is λ(ψ) in ψ k Hessian formation Approximation of Δ k It is the trust region of the k-th iteration, λ k Is λ(ψ) in ψ k The value at;

[0026] Solving the quadratic approximation model yields the identification parameters M. ps1 M ps2 …M psn and R L1 ,R L2 …R Ln The final solution.

[0027] Preferably, in the first objective function F(ψ) and the second objective function λ(ψ), I pi The expression for (ψ) is:

[0028]

[0029] Among them, U in(i) This indicates that the output resonant frequency of the high-frequency inverter 2 is f. i voltage u ini The root mean square value, Z r(i) =Z rRe(i) +Z rIm(i) Z represents the system input impedance. rRe(i) Z rIm(i) These represent the real and imaginary parts of the system input impedance, respectively.

[0030] In the second objective function λ(ψ), tanθ i The expression for (ψ) is:

[0031] tanθ i (ψ)=Z rIm(i) / Z rRe(i) .

[0032] Preferably, the primary-side energy emission module 3 includes a primary-side magnetic energy emission coil L. p Primary magnetic energy transmitting coil L p The equivalent series resistance is R p Secondary magnetic energy pickup coil L si The equivalent series resistance is R si Secondary magnetic energy pickup coil L si The secondary compensation capacitor connected is denoted as C. si Then we have:

[0033]

[0034]

[0035] Where, ω i To be with f i The corresponding angular frequency.

[0036] Preferably, the NSGA II algorithm is used to solve the first optimization problem, and the TrustRegion algorithm is used to solve the quadratic approximation model.

[0037] This invention also provides a control method for a multi-frequency, multi-load WPT system based on parameter identification, the key of which includes the following steps:

[0038] S1. Collect the output current of the high-frequency inverter 2, i.e., the transmitter-side current;

[0039] S2. Perform FFT analysis on the acquired transmitter-side current to obtain the effective value I of the transmitter-side current at different frequencies. p1mea ,I p2mea …Ipnmea ;

[0040] S3, according to I p1mea ,I p2mea …I pnmea Calculate the identification parameter M ps1 M ps2 …M psn and R L1 ,R L2 …R Ln ;

[0041] S4. Based on the identification parameter M ps1 M ps2 …M psn and R L1 ,R L2 …R Ln Change the operating state of the high-frequency inverter 2 to adapt it to the current load requirements.

[0042] Furthermore, step S3 specifically includes the following steps:

[0043] S31, Identify parameter M ps1 M ps2 …M psn and R L1 ,R L2 …R Ln As the optimization variable for the multi-objective optimization problem, the effective value of the primary current I... p1 ,I p2 …I pn The measured value I p1mea ,I p2mea …I pnmea Compared with the theoretically calculated value I p1 (ψ),I p2 (ψ)…I pn By subtracting (ψ) from each other, we construct the first objective function F(ψ) and the first optimization problem:

[0044] P1:

[0045] stM psiL <M psi <M psiH ,R LiL <R Li <R LiH i = 1, 2, ..., n

[0046] Among them, M psiL M psiH M respectively psi The upper and lower limits of R LiL R LiH R respectively LiThe upper and lower limits;

[0047] S32. Solve the first optimization problem to obtain the identification parameters M. ps1 M ps2 …M psn and R L1 ,R L2 …R Ln The initial solution.

[0048] Step S2 further includes:

[0049] FFT analysis was performed on the acquired transmitter-side current to obtain the phase differences θ1, θ2…θ between the transmitter-side current and voltage at different frequencies. n tangent value tanθ 1mea ,tanθ 2mea …tanθ nmea ;

[0050] Step S3 further includes the following steps:

[0051] S33, Identify parameter M ps1 M ps2 …M psn and R L1 ,R L2 …R Ln Using the initial solution as the initial value, the effective value of the primary current I p1 ,I p2 …I pn The measured value I p1mea ,I p2mea …I pnmea Compared with the theoretically calculated value I p1 (ψ),I p2 (ψ)…I pn (ψ) are calculated separately, and the tangent of the phase difference between the primary voltage and the current is measured as tanθ. 1mea ,tanθ 2mea …tanθ nmea By subtracting the theoretically calculated values ​​tanθ1(ψ), tanθ1(ψ)…tanθ1(ψ) respectively, a second objective function λ(ψ) is constructed, and a second optimization problem is established:

[0052] P2:

[0053] S34. Perform a second-order Taylor expansion on the second objective function λ(ψ) to obtain a quadratic approximation model:

[0054]

[0055] st||d||≤Δ k

[0056] Where d=ψ-ψ k Indicates the trial step size, g k The second objective function λ(ψ) at the current iteration point ψ k gradient at point B k Is λ(ψ) in ψ k Hessian formation Approximation of Δ k It is the trust region of the k-th iteration, λ k Is λ(ψ) in ψ k The value at;

[0057] S35. Solve the quadratic approximation model to obtain the identification parameter M. ps1 M ps2 …M psn and R L1 ,R L2 …R Ln The final solution.

[0058] This invention provides a multi-frequency, multi-load WPT system and its control method based on parameter identification. Compared with traditional parameter identification methods for WPT systems, its advantages are as follows:

[0059] 1) By accurately measuring the primary current directly, the magnitude of the load and mutual inductance parameters can be calculated, thereby enabling the identification of load and mutual inductance parameters in a multi-frequency, multi-load WPT system without affecting the normal operation of the system. In other words, it enables the simultaneous identification of multi-frequency, multi-load system parameters and power transmission.

[0060] 2) Frequency and power are easy to adjust and can be adjusted according to requirements to meet load needs;

[0061] 3) Load and mutual inductance parameter identification is based on NSGAⅡ algorithm and TrustRegion algorithm, with high identification accuracy and no need for additional control devices;

[0062] 4) It solves the problems of traditional parameter identification methods, which can only identify a single load or mutual inductance parameter, resulting in inaccurate results and complex circuits that make detection difficult. Attached Figure Description

[0063] Figure 1 This is an architecture diagram of the multi-frequency, multi-load WPT system based on parameter identification provided in an embodiment of the present invention;

[0064] Figure 2 This is a flowchart of the parameter identification algorithm provided in an embodiment of the present invention;

[0065] Figure 3 This is a flowchart of parameter identification when n=2 provided in an embodiment of the present invention;

[0066] Figure 4 This is an FFT plot of the primary-side energy emission module current at a frequency of 20kHz provided in an embodiment of the present invention;

[0067] Figure 5 This is an FFT analysis result diagram of the primary-side energy emission module current at various frequencies provided in the embodiments of the present invention;

[0068] Figure 6 This is a statistical result diagram of parameter identification under different loads provided in the embodiments of the present invention, where (a), (b), (c), and (d) correspond to parameter R respectively. L1 M ps1 R L2 M ps2 .

[0069] Figure reference numerals: 1-DC power supply; 2-high frequency inverter; 3-primary side energy transmitting module; 4-secondary side energy receiving module; 5-system load; 6-primary side signal control and processing module. Detailed Implementation

[0070] The embodiments of the present invention are described in detail below with reference to the accompanying drawings. The embodiments are given for illustrative purposes only and should not be construed as limiting the present invention. The accompanying drawings are for reference and illustration only and do not constitute a limitation on the scope of patent protection of the present invention, because many changes can be made to the present invention without departing from the spirit and scope of the present invention.

[0071] This invention provides a multi-frequency, multi-load WPT system based on parameter identification, such as... Figure 1 As shown, it includes a DC power supply 1, a high-frequency inverter 2, a primary-side energy transmitting module 3, a secondary-side energy receiving module 4, a system load 5, and a primary-side signal control and processing module 6.

[0072] Primary-side energy emission module 3 includes primary-side magnetic energy emission coil L p Primary magnetic energy transmitting coil L p The equivalent series resistance is R p .

[0073] The secondary side energy receiving module 4 includes n ≥ 2 natural resonant frequencies f1, f2, ... f n The power receiving circuit comprises a secondary magnetic energy pickup coil and a secondary compensation capacitor connected in series. The i-th power receiving circuit includes a secondary magnetic energy pickup coil L connected in sequence. si The internal resistance R of the magnetic energy pickup coil si and secondary side compensation capacitor C si , i = 1, 2…n. Secondary magnetic energy pickup coil L siThe equivalent series resistance is R si According to the resonance relationship, the secondary magnetic energy pickup coil L in the secondary energy receiving module... si and secondary side compensation capacitor C si satisfy

[0074] System load 5 includes a load connected in series with n secondary magnetic energy pickup coils, including R L1 ,R L2 …R Ln .

[0075] The primary-side signal control and processing module 6 includes a current detection circuit, a DSP module, and a composite modulation PWM control circuit connected in sequence.

[0076] The composite modulation PWM control circuit is used to output n composite modulation waves of different frequencies according to n different inherent resonant frequencies of the power receiving circuit to drive the high-frequency inverter 2 to work.

[0077] The current detection circuit is used to acquire the transmitter-side current; the DSP module is used to perform FFT analysis on the transmitter-side current acquired by the current detection circuit to obtain the effective value I of the transmitter-side current at different frequencies. p1mea ,I p2mea …I pnmea , and according to I p1mea ,I p2mea …I pnmea Calculate the identification parameter M ps1 M ps2 …M psn and R L1 ,R L2 …R Ln M psi This indicates that the primary side energy emission module 3 and the i-th secondary side magnetic energy pickup coil L are connected. si Mutual intuition between them, R Li This represents the load in system load 5 that is connected to the i-th secondary magnetic energy pickup coil, where i = 1, 2, ..., n;

[0078] The DSP module is also used to identify parameters M ps1 M ps2 …M psn and R L1 ,R L2 …R Ln The control circuit of composite modulation PWM outputs a new composite modulation wave to change the operating state of the high-frequency inverter 2 to adapt to the current load requirements.

[0079] DSP module according to I p1mea ,I p2mea …I pnmea Calculate the identification parameter M ps1M ps2 …M psn and R L1 ,R L2 …R Ln The specific process is as follows:

[0080] Identification parameter M ps1 M ps2 …M psn and R L1 ,R L2 …R Ln As the optimization variable for the multi-objective optimization problem, the effective value of the primary current I... p1 ,I p2 …I pn The measured value I p1mea ,I p2mea …I pnmea Compared with the theoretically calculated value I p1 (ψ),I p2 (ψ)…I pn By subtracting (ψ) from each other, we construct the first objective function F(ψ) and the first optimization problem:

[0081] P1:

[0082] stM psiL <M psi <M psiH ,R LiL <R Li <R LiH i = 1, 2, ..., n

[0083] Among them, M psiL M psiH M respectively psi The upper and lower limits of R LiL R LiH R respectively Li The upper and lower limits;

[0084] This example uses the NSGA II algorithm to solve the first optimization problem and obtain the identification parameters M. ps1 M ps2 …M psn and R L1 ,R L2 …R Ln The initial solution is obtained, and the DSP module can then adjust the high-frequency inverter based on this initial value.

[0085] Although the identification parameter M has been obtained ps1 M ps2 …M psn and R L1 ,R L2 …RLn However, its initial solution is not accurate enough. To make the calculation more accurate, the DSP module in this example also performs FFT analysis on the transmitter current collected by the current detection circuit to obtain the phase difference θ1, θ2…θ between the transmitter current and voltage at different frequencies. n tangent value tanθ 1mea ,tanθ 2mea …tanθ nmea And based on the initial solution already obtained and I p1mea ,I p2mea …I pnmea and tanθ 1mea ,tanθ 2mea …tanθ nmea Calculate the identification parameter M ps1 M ps2 …M psn and R L1 ,R L2 …R Ln The final solution is obtained. The DSP module controls the high-frequency inverter 2 based on this final solution, making the system control more precise.

[0086] Having obtained the initial solution, the DSP module further... p1mea ,I p2mea …I pnmea and tanθ 1mea ,tanθ 2mea …tanθ nmea Calculate the identification parameter M ps1 M ps2 …M psn and R L1 ,R L2 …R Ln The specific process is as follows:

[0087] Identification parameter M ps1 M ps2 …M psn and R L1 ,R L2 …R Ln Using the initial solution as the initial value, the effective value of the primary current I p1 ,I p2 …I pn The measured value I p1mea ,I p2mea …I pnmea Compared with the theoretically calculated value I p1 (ψ),I p2 (ψ)…I pn (ψ) are calculated separately, and the tangent of the phase difference between the primary voltage and the current is measured as tanθ. 1mea ,tanθ2mea …tanθ nmea By subtracting the theoretically calculated values ​​tanθ1(ψ), tanθ1(ψ)…tanθ1(ψ) respectively, a second objective function λ(ψ) is constructed, and a second optimization problem is established:

[0088] P2:

[0089] To simplify the calculation, the second objective function λ(ψ) is expanded using Taylor second order, resulting in a quadratic approximation model:

[0090]

[0091] st||d||≤Δ k

[0092] Where d=ψ-ψ k Indicates the trial step size, g k The second objective function λ(ψ) at the current iteration point ψ k gradient at point B k Is λ(ψ) in ψ k Hessian formation Approximation of Δ k It is the trust region of the k-th iteration, λ k Is λ(ψ) in ψ k The value at;

[0093] The TrustRegion algorithm is used to solve the quadratic approximation model to obtain the identification parameters M. ps1 M ps2 …M psn and R L1 ,R L2 …R Ln The final solution. To measure the quadratic model m... k (d k ) and objective function λ(ψ) k +d k The degree of approximation of ), defining the ratio. The optimal solution for the optimization variable ψ is obtained through iterative approximation.

[0094] The overall algorithm flow is as follows: Figure 2 As shown.

[0095] In the first objective function F(ψ) and the second objective function λ(ψ), I pi The expression for (ψ) is:

[0096]

[0097] Among them, U in(i) This indicates that the output resonant frequency of high-frequency inverter 2 is f.i voltage u ini The root mean square value, Z r(i) =Z rRe(i) +Z rIm(i) Z represents the system input impedance. rRe(i) Z rIm(i) These represent the real and imaginary parts of the system input impedance, respectively.

[0098] In the second objective function λ(ψ), tanθ i The expression for (ψ) is:

[0099] tanθ i (ψ)=Z rIm(i) / Z rRe(i) .

[0100] according to Figure 1 The system shown can be derived from Kirchhoff's voltage law as follows:

[0101]

[0102]

[0103] Where, ω i To be with f i The corresponding angular frequency.

[0104] Corresponding to the above system, the present invention also provides a control method for a multi-frequency, multi-load WPT system based on parameter identification, referencing... Figure 3 The flowchart shown below illustrates the method for n=2, and includes the following steps:

[0105] S1. Collect the output current of high-frequency inverter 2, i.e., the transmitter-side current;

[0106] S2. Perform FFT analysis on the acquired transmitter-side current to obtain the effective value I of the transmitter-side current at different frequencies. p1mea ,I p2mea …I pnmea ;

[0107] S3, according to I p1mea ,I p2mea …I pnmea Calculate the identification parameter M ps1 M ps2 …M psn and R L1 ,R L2 …R Ln ;

[0108] S4. Based on the identification parameter M ps1 M ps2 …M psn and R L1,R L2 …R Ln Change the operating state of high-frequency inverter 2 to adapt it to the current load requirements.

[0109] Step S3 specifically includes the following steps:

[0110] S31, Identify parameter M ps1 M ps2 …M psn and R L1 ,R L2 …R Ln As the optimization variable for the multi-objective optimization problem, the effective value of the primary current I... p1 ,I p2 …I pn The measured value I p1mea ,I p2mea …I pnmea Compared with the theoretically calculated value I p1 (ψ),I p2 (ψ)…I pn By subtracting (ψ) from each other, we construct the first objective function F(ψ) and the first optimization problem:

[0111] P1:

[0112] stM psiL <M psi <M psiH ,R LiL <R Li <R LiH i = 1, 2, ..., n

[0113] Among them, M psiL M psiH M respectively psi The upper and lower limits of R LiL R LiH R respectively Li The upper and lower limits;

[0114] S32. Solve the first optimization problem to obtain the identification parameters M. ps1 M ps2 …M psn and R L1 ,R L2 …R Ln The initial solution.

[0115] Similarly, since the initial solution's identification accuracy is not high enough, step S2 further includes:

[0116] FFT analysis was performed on the acquired transmitter-side current to obtain the phase differences θ1, θ2…θ between the transmitter-side current and voltage at different frequencies.n tangent value tanθ 1mea ,tanθ 2mea …tanθ nmea ;

[0117] Therefore, step S3 also includes the following steps:

[0118] S33, Identify parameter M ps1 M ps2 …M psn and R L1 ,R L2 …R Ln Using the initial solution as the initial value, the effective value of the primary current I p1 ,I p2 …I pn The measured value I p1mea ,I p2mea …I pnmea Compared with the theoretically calculated value I p1 (ψ),I p2 (ψ)…I pn (ψ) are calculated separately, and the tangent of the phase difference between the primary voltage and the current is measured as tanθ. 1mea ,tanθ 2mea …tanθ nmea By subtracting the theoretically calculated values ​​tanθ1(ψ), tanθ1(ψ)…tanθ1(ψ) respectively, a second objective function λ(ψ) is constructed, and a second optimization problem is established:

[0119] P2:

[0120] S34. Perform a second-order Taylor expansion on the second objective function λ(ψ) to obtain a quadratic approximation model:

[0121]

[0122] st||d||≤Δ k

[0123] Where d=ψ-ψ k Indicates the trial step size, g k The second objective function λ(ψ) at the current iteration point ψ k gradient at point B k Is λ(ψ) in ψ k Hessian formation Approximation of Δ k It is the trust region of the k-th iteration, λ k Is λ(ψ) in ψ k The value at m k (d) is the value of the quadratic approximation model when the step size is d in the kth iteration;

[0124] S35. Solve the quadratic approximation model to obtain the identification parameters M. ps1 M ps2 …M psn and R L1 ,R L2 …R Ln The final solution.

[0125] Other aspects of this method are consistent with the system and will not be repeated in this embodiment.

[0126] In summary, the multi-frequency, multi-load WPT system and its control method based on parameter identification provided by this invention have the following advantages compared to traditional parameter identification methods for WPT systems:

[0127] 1) By accurately measuring the primary current directly, the magnitude of the load and mutual inductance parameters can be calculated, thereby enabling the identification of load and mutual inductance parameters in a multi-frequency, multi-load WPT system without affecting the normal operation of the system. In other words, it enables the simultaneous identification of multi-frequency, multi-load system parameters and power transmission.

[0128] 2) Frequency and power are easy to adjust and can be adjusted according to requirements to meet load needs;

[0129] 3) Load and mutual inductance parameter identification is based on NSGAⅡ algorithm and TrustRegion algorithm, with high identification accuracy and no need for additional control devices;

[0130] 4) It solves the problems of traditional parameter identification methods, which can only identify a single load or mutual inductance parameter, resulting in inaccurate results and complex circuits that make detection difficult.

[0131] The parameter identification effect of the present invention embodiment is verified below with n=2.

[0132] Primary magnetic energy transmitting coil L p The inductance is 29.1 μH, and the internal resistance R of the transmitting coil is... p The resistance is 0.05Ω, and the secondary magnetic energy pickup coil L... s1 L s2 The inductance values ​​are 97.6μH and 97.6μH respectively, and the internal resistance R is... s1 R s2 The resistance is 0.08Ω, and the secondary compensation capacitor C... s1 C s2 The capacitor values ​​are 648.83nF and 35.921nF respectively, the DC power supply voltage is 48V, and the voltage modulation a1 and a2 are both set to 0.23 at each frequency. This causes the two receiving circuits of the secondary energy receiving module 4 to resonate at the system operating frequencies of 20kHz and 85kHz respectively. Figure 4 This is an FFT plot of the primary-side energy emission module current at a frequency of 20kHz under certain mutual inductance and load parameters according to the present invention. Figure 5 This is an FFT analysis result of the primary-side energy emission module current at various frequencies under certain mutual inductance and load parameters of the present invention.

[0133] The high-frequency inverter 2 outputs voltages u at frequencies of 20kHz and 85kHz, respectively, for f1 and f2. in1 and u in2 The root mean square value is

[0134] According to Kirchhoff's voltage law, the input impedance and the expressions for its real and imaginary parts of a dual-frequency, dual-load WPT system can be obtained, as follows:

[0135] Z r(i) =Z rRe(i) +Z rIm(i)

[0136]

[0137]

[0138] Where i = 1, 2.

[0139] Based on the above data, the primary voltage can be calculated as follows:

[0140]

[0141] The system's operating angular frequency is:

[0142] ω1=2πf1=2π×20000=40000πHz

[0143] ω2=2πf2=2π×85000=170000πHz

[0144] Combining the calculation formula analyzed above, we have:

[0145]

[0146] in:

[0147]

[0148]

[0149]

[0150]

[0151] Furthermore, we can obtain:

[0152]

[0153] Where, tanθ i (ψ)=Z rIm(i) / Z rRe(i) , i = 1, 2.

[0154] The identification algorithm can be used to obtain:

[0155]

[0156] The error is:

[0157]

[0158] It can be seen that the errors of the calculated load and mutual inductance parameters are all within 3% of the engineering error, with an average error of about 1.42%, which shows high parameter identification accuracy.

[0159] In addition, this embodiment also conducted multiple experiments by replacing different loads, and the statistical results are as follows: Figure 6 Where (a), (b), (c), and (d) correspond to parameters R respectively. L1 M ps1 R L2 M ps2 As can be seen, the identification errors of the system parameters are all within an acceptable range, which fully verifies the feasibility of the present invention.

[0160] The above embodiments are preferred embodiments of the present invention, but the embodiments of the present invention are not limited to the above embodiments. Any changes, modifications, substitutions, combinations, or simplifications made without departing from the spirit and principle of the present invention shall be considered equivalent substitutions and shall be included within the protection scope of the present invention.

Claims

1. A multi-frequency multi-load WPT system based on parameter identification, characterized by, The system comprises a direct current power supply (1), a high-frequency inverter (2), a primary energy transmitting module (3), a secondary energy receiving module (4), a system load (5) and a primary signal control and processing module (6). The secondary side energy receiving module (4) comprises n≥2 electric energy receiving circuits with inherent resonance frequencies f1, f2, … f n n respectively, each of the electric energy receiving circuits comprising a secondary side magnetic energy pickup coil and a secondary side compensation capacitor connected in series. The primary signal control and processing module (6) comprises sequentially connected current detection circuit, DSP module and composite modulation PWM control circuit. The composite modulation PWM control circuit is used for outputting n composite modulation waves of different frequencies according to n different inherent resonance frequencies of the power receiving circuit to drive the high-frequency inverter (2) to work. The current detection circuit is used for collecting the transmitting side current; the DSP module is used for performing FFT analysis on the transmitting side current collected by the current detection circuit to obtain the current effective value I of the transmitting side under different frequencies p1mea ,I p2mea …I pnmea And according to I p1mea ,I p2mea …I pnmea Calculate the identification parameter M ps1 ,M ps2 …M psn And R L1 ,R L2 …R Ln , M psi Indicates the mutual inductance between the primary energy transmitting module (3) and the i-th secondary side magnetic energy pickup coil L si , R Li Indicates the load connected with the i-th secondary side magnetic energy pickup coil in the system load (5), i=1,2…n; The DSP module is also used to determine the identification parameter M ps1 ,M ps2 …M psn and R L1 ,R L2 …R Ln control the composite modulation PWM control circuit to output a new composite modulation wave to change the working state of the high-frequency inverter (2) to adapt to the current load demand; The DSP module calculates the recognition parameter M p1mea , p2mea … pnmea The specific process of calculating the recognition parameter M ps1 , ps2 …M psn and R L1 , L2 …R Ln is as follows: The identification parameter M ps1 M ps2 …M psn and R L1 R L2 …R Ln As the optimization variable ψ of the multi-objective optimization problem, the measured value I p1 I p2 I pn of the original side current effective value I p1mea I p2mea I pnmea and the theoretical calculation value I p1 (ψ), I p2 (ψ)…I pn (ψ) are respectively subtracted, the first objective function F(ψ) is constructed, and the first optimization problem is constructed: P1： ， s.t. M psiL <M psi <M psiH , R LiL <R Li <R LiH , i=1,2…n, Among them, M psiL M psiH M respectively psi The upper and lower limits of R LiL R LiH R respectively Li The upper and lower limits; solving the first optimization problem to obtain identified parameters M ps1 ,M ps2 …M psn and R L1 ,R L2 …R Ln initial solution.

2. The parameter identification based multi-frequency multi-load WPT system of claim 1, wherein, The DSP module is also used to perform FFT analysis on the transmitter current acquired by the current detection circuit to obtain the phase difference θ1, θ2…θ between the transmitter current and voltage at different frequencies. n tangent value tanθ 1mea ,tanθ 2mea …tanθ nmea It is also used according to I p1mea ,I p2mea …I pnmea tanθ 1mea ,tanθ 2mea …tanθ nmea and identification parameter M ps1 M ps2 …M psn and R L1 ,R L2 …R Ln The initial solution is used to calculate the identification parameters M. ps1 M ps2 …M psn and R L1 ,R L2 …R Ln The final solution.

3. The parameter identification based multi-frequency multi-load WPT system of claim 2, wherein, The DSP module is based on I p1mea ,I p2mea …I pnmea tanθ 1mea ,tanθ 2mea …tanθ nmea and identification parameter M ps1 M ps2 …M psn and R L1 ,R L2 …R Ln The initial solution is used to calculate the identification parameters M. ps1 M ps2 …M psn and R L1 ,R L2 …R Ln The specific process of the final solution is as follows: Identification parameter M ps1 M ps2 …M psn and R L1 ,R L2 …R Ln Using the initial solution as the initial value, the effective value of the primary current I p1 ,I p2 …I pn The measured value I p1mea ,I p2mea …I pnmea Compared with the theoretically calculated value I p1 (ψ),I p2 (ψ)…I pn (ψ) are calculated separately, and the tangent of the phase difference between the primary voltage and the current is measured as tanθ. 1mea ,tanθ 2mea …tanθ nmea By subtracting the theoretically calculated values ​​tanθ1(ψ), tanθ1(ψ)…tanθ1(ψ) respectively, a second objective function λ(ψ) is constructed, and a second optimization problem is established: P2： ， The second objective function λ(ψ) is Taylor expanded to the second order to obtain a quadratic approximation model: ， ， where d = ψ - ψ k denotes the trial step, g k is the gradient of the second objective function λ(ψ) at the current iteration point ψ k B k is an approximation of the Hessian matrix ∇ k λ(ψ) at ψ 2 Δ k is the trust region at the k-th iteration, is the value of λ(ψ) at ψ k ; solving the quadratic approximation model to obtain the identification parameters M ps1 ,M ps2 …M psn and R L1 ,R L2 …R Ln the final solution.

4. The parameter identification based multi-frequency multi-load WPT system of claim 3, wherein: In the first objective function F(ψ) and the second objective function λ(ψ), I pi The expression of F(ψ) is: ， wherein denotes the resonance frequency of the high-frequency inverter (2) output voltage u i of the high-frequency inverter (2) output voltage u ini of the high-frequency inverter (2) output voltage u denotes the system input impedance, , denote the real and imaginary parts of the system input impedance, respectively; In the second objective function λ(ψ), tan θ i The expression for λ(ψ) is: 。 5. The parameter identification based multi-frequency multi-load WPT system of claim 4, wherein: The primary energy emission module (3) comprises a primary magnetic energy emission coil L p , the equivalent series resistance of the primary magnetic energy emission coil L p is R p , the equivalent series resistance of the secondary magnetic energy pickup coil L si is R si , and the secondary compensation capacitor connected to the secondary magnetic energy pickup coil L si is C si , then: ， ， wherein f is the frequency of the signal i corresponding angular frequency.

6. The parameter identification based multi-frequency multi-load WPT system of claim 3, wherein: The first optimization problem is solved by using the NSGA II algorithm, and the quadratic approximation model is solved by using the Trust Region algorithm.

7. The control method of the multi-frequency multi-load WPT system based on parameter identification according to any one of claims 1-6, characterized in that, The method comprises the following steps: S1, collecting the output current of the high-frequency inverter (2), i.e. the transmitting side current; S2, FFT analysis is performed on the collected transmitting-side current to obtain the effective value I of the transmitting-side current at different frequencies p1mea ,I p2mea …I pnmea ; S3, according to I p1mea ,I p2mea …I pnmea Computing the recognition parameter M ps1 ,M ps2 …M psn and R L1 ,R L2 …R Ln ; S4. Adapt the operating state of the high frequency inverter (2) to the current load demand based on the identified parameter M ps1 M ps2 …M psn and R L1 R L2 …R Ln changing the operating state of the high frequency inverter (2) to adapt it to the current load demand The step S3 specifically comprises the following steps: S31, identify parameters M ps1 ,M ps2 …M psn and R L1 ,R L2 …R Ln As the optimization variable ψ of the multi-objective optimization problem, the effective value I p1 ,I p2 …I pn of the original side current is taken as the measured value I p1mea ,I p2mea …I pnmea and the theoretically calculated value I p1 (ψ), I p2 (ψ)…I pn (ψ) are respectively subtracted, a first objective function F(ψ) is constructed, and a first optimization problem is constructed: P1： ， s.t. M psiL <M psi <M psiH , R LiL <R Li <R LiH , i=1,2…n, Among them, M psiL M psiH M respectively psi The upper and lower limits of R LiL R LiH R respectively Li The upper and lower limits; S32, solve the first optimization problem to obtain identification parameter M ps1 ,M ps2 …M psn and R L1 ,R L2 …R Ln initial solution.

8. The control method of a parameter identification based multi-frequency multi-load WPT system according to claim 7, wherein, The step S2 further comprises the following steps: The collected transmitting side current is subjected to FFT analysis to obtain phase difference θ1, θ2... θ between the current and voltage of the transmitting side at different frequencies n The tangent value tanθ 1mea of the phase difference θ1, θ2... θ 2mea The tangent value tanθ nmea of the phase difference θ1, θ2... θ The step S3 further comprises the following steps: S33, the identification parameter M ps1 M ps2 …M psn and R L1 R L2 …R Ln the initial solution as the initial value, the primary current effective value I p1 I p2 I pn the measured value I p1mea I p2mea I pnmea and the theoretical calculation value I p1 (ψ), I p2 (ψ)…I pn (ψ) respectively, and the measured value tanθ 1mea tanθ 2mea …tanθ nmea and the theoretical calculation value tanθ1(ψ), tanθ1(ψ)…tanθ1(ψ) respectively, construct the second objective function λ(ψ), and construct the second optimization problem: P2： ， S34, Taylor expanding the second objective function λ(ψ) to the second order to obtain a quadratic approximation model: ， ， where d = ψ - ψ k denotes the trial step, g k is the gradient of the second objective function λ(ψ) at the current iteration point ψ k B k is an approximation of the Hessian matrix ∇ k λ(ψ) at ψ 2 Δ k is the trust region at the k-th iteration, is the value of λ(ψ) at ψ k ; S35, solving the twice approximate model to obtain identification parameter M ps1 ,M ps2 …M psn and R L1 ,R L2 …R Ln final solution.

Images