An ultrasonic machining system compensation network parameter optimization method

Abstract

The application discloses an ultrasonic machining system compensation network parameter optimization method, first, the compensation network of the non-contact electric energy transmission system is designed, so that the non-contact electric energy transmission system has better anti-interference ability and impedance characteristics, and larger power can be output under low voltage. Secondly, the transmission performance evaluation function and the evaluation function of the quality factor, the admittance circle radius and the coupling coefficient are respectively proposed from the transmission performance and the impedance characteristic angle, and the parameter optimization model under the variable load in the ultrasonic machining process is constructed. Finally, the compensation element parameters of the ultrasonic vibrator under the variable load are obtained by using the multi-objective genetic algorithm. The application solves the problem that the transmission efficiency and the output power of the traditional basic topology cannot reach a high level at the same time, and influences the machining performance of the system, improves the overall transmission performance and anti-interference of the system under the variable load, and enhances the load capacity of the system.

Description

Technical Field

[0001] This invention belongs to the field of non-contact energy transmission, specifically relating to a method for optimizing compensation network parameters in an ultrasonic processing system. Background Technology

[0002] In rotary ultrasonic machining (RUM), the application of non-contact energy transfer theory solves the limitation of traditional contact power supply systems on rotation speed, but it also brings the problem of poor transmission performance. Especially in the actual ultrasonic machining and ultrasonic transducer design and manufacturing process, the impedance characteristics of the transducer itself will always change to some extent. In this case, the non-contact energy transfer system (CETS) is difficult to maintain high transmission efficiency and good impedance characteristics, which affects machining performance and workpiece quality.

[0003] In ultrasonic high-speed machining, if a power supply method with direct contact between the brush and the slip ring is used, it will not only cause a series of problems such as high frictional loss, high heat generation, and easy arcing between the brush and the slip ring, limiting the high-speed rotation of ultrasonic machining and posing serious safety hazards to the actual machining process, but the piezoelectric transducer will also undergo performance changes due to the heat generated by friction over a long period of time, resulting in a shift in the resonant frequency and affecting the actual machining effect. At the same time, it will cause severe wear on the tool head. Therefore, the theory of non-contact power transfer is particularly important for maintaining safe and stable high-speed rotation in ultrasonic machining. This transmission method can separate the primary and secondary sides of the transformer, avoiding heat generation due to friction, without affecting the performance of the piezoelectric transducer or limiting the rotational speed.

[0004] However, the inductance of the primary and secondary coils affects the transmission performance of CETS. Several basic two-sided topologies have been developed for compensation network structures, such as primary-secondary series (SS), primary-secondary series parallel (SP), primary-secondary parallel series (PS), and primary-secondary parallel (PP). Regarding the parameters of the compensation components, existing compensation network structures only optimize for the single resonant frequency of the ultrasonic transducer under no-load conditions. However, in reality, the ultrasonic transducer will inevitably experience different loads during manufacturing, or dynamic component parameters may deviate to some extent during the design and fabrication process, leading to resonant frequency shifts and deteriorated impedance characteristics. This reduces the performance of the compensation network and affects the actual manufacturing results. Summary of the Invention

[0005] To address the aforementioned problems, this invention proposes a method for optimizing the parameters of a compensation network in an ultrasonic processing system. This method optimizes the structure and parameters of the traditional compensation network, enabling the system's transmission efficiency and impedance characteristics to remain at a high level even when the dynamic components of the ultrasonic transducer experience a certain range of deviations, thereby improving the transmission performance and anti-interference capability of the CETS (Cyclic Electrification System).

[0006] This invention first designs a primary-secondary side compensation network with an LC-CL structure, which gives the non-contact energy transmission system better anti-interference capability and impedance characteristics, and enables it to output greater power even at low voltage, effectively improving the transmission performance of the processing system. Secondly, based on the fluctuation of equivalent dynamic element parameters during actual processing and ultrasonic transducer fabrication, a variable load parameter optimization model is proposed, which significantly improves the impedance characteristics of the LC-CL structure compared to the traditional SS primary-secondary side series compensation structure, effectively enhancing its load-carrying capacity.

[0007] The present invention includes the following specific steps:

[0008] S1. Compensation network design for contactless power transmission system:

[0009] The compensation form of the primary side of the compensation network is determined, second-order CL compensation is performed on the secondary side, and a capacitor is selected as the primary side series compensation element.

[0010] Based on the original compensation structure, a capacitor element is connected in parallel on the primary side and an inductor element is connected in series to form an LC-CL compensation structure.

[0011] S2, Optimization model for compensation element parameters:

[0012] S2.1 Transmission Performance Evaluation

[0013] The efficiency performance evaluation function, based on the transmission efficiency formula of the compensation structure, yields the transmission efficiency η under different load conditions and different compensation network parameters, ensuring that the system achieves a high level of transmission efficiency with minimal efficiency deviation under various load conditions. The equivalent circuit parameters of the ultrasonic transducer under n loads are measured using an impedance analyzer, and the above measurement process is repeated m times. In each measurement, p loads are randomly selected from the n loads for transmission efficiency calculation, and the weight w for each measurement is determined using the Coefficient of Variation method.

[0014] The specific calculation process is as follows: the average system transmission efficiency A during each measurement. i and standard deviation S i for:

[0015]

[0016] Where η(i,j) represents the system transmission efficiency under the j-th random load during the i-th measurement.

[0017] The standard deviation describes the dispersion of the values, thus yielding the coefficient of variation V of the system transmission efficiency for each measurement. i :

[0018]

[0019] After standardization, the weight w of the i-th measurement is obtained. i :

[0020]

[0021] The efficiency evaluation function F1, which determines the optimal transmission efficiency or output power performance with the smallest deviation, is:

[0022]

[0023] Among them, L ps For the primary-side series inductance, C pp The primary side is connected in parallel with capacitor C. sp For the secondary side parallel capacitor, L ss It is a series inductor on the secondary side.

[0024] According to the F1 evaluation function, the greater the fluctuation of the transmission efficiency value between different load conditions in the m measurement groups, the greater the weight of that group. In order to effectively reduce the relative deviation, this invention multiplies the obtained weight by the minimum efficiency value in each measurement group, and the largest value of the calculation result is the evaluation result with the best transmission efficiency and the smallest relative deviation.

[0025] S2.2 Impedance Performance Evaluation

[0026] Multiple evaluation metrics are obtained using the admittance circle diagram of the CETS circuit, and the compensation network parameters are optimized to improve the load-carrying capacity and adaptability of the ultrasonic machining system during actual machining. Based on some characteristic points and eigenvalues ​​of the admittance circle of the CETS circuit, several key indicators for the ultrasonic machining system are selected as evaluation functions in the selection process of compensation network parameters:

[0027] S2.2.1 Mechanical quality factor Q of the system circuit m Q is an important technical indicator for evaluating rotary ultrasonic machining systems, characterizing the energy consumed by a piezoelectric material to overcome internal friction at resonance. The lower the mechanical loss, the higher the Q value. m The larger the value, the more its calculation expression can be represented as:

[0028]

[0029] In the formula, f1 and f2 are the half-power points of the circuit system, f r It is the resonant frequency.

[0030] The mean mechanical quality factor of the system circuit under n load conditions is used as the quality factor evaluation function F2:

[0031]

[0032] In the formula, Q m (i) represents the mechanical quality factor under the i-th load condition.

[0033] S2.2.2 To increase the admittance circle radius of the system circuit, the average value of the admittance circle radius under n load conditions is selected as the admittance radius evaluation function F3:

[0034]

[0035] In the formula, R0(i) represents the resonant resistance of the system circuit under the i-th load condition.

[0036] S2.2.3, The electromechanical coupling coefficient is a parameter used to represent the coupling relationship between mechanical energy and electrical energy. It reflects the amount of mechanical energy that can be extracted from the total energy of the piezoelectric transducer to provide driving power, and it is also an important factor determining the bandwidth. The coupling coefficient k is related to the forward and reverse resonant frequencies f. r ,f a The calculation relationship between them is:

[0037]

[0038] Therefore, in order to improve the electromechanical coupling performance of the system, increase the system bandwidth, and enhance the system's anti-interference capability during the processing, the average electromechanical coupling coefficient of the system under n load conditions is used as the electromechanical coupling evaluation function F4:

[0039]

[0040] In the formula, k(i) represents the electromechanical coupling coefficient under the i-th load condition.

[0041] S2.2.4 Selecting the resonant frequency f of the CETS circuit under different load conditions r The resonant frequency f of the ultrasonic transducer under actual conditions t The average of the differences is used as the frequency deviation constraint function H:

[0042]

[0043] In the formula, f r (i) represents the resonant frequency of the rotary ultrasonic machining system circuit under the i-th load condition, f t (i) represents the resonant frequency of the ultrasonic transducer under the i-th load condition.

[0044] S2.3 Performance Evaluation Function

[0045] Combining the above four evaluation functions and one constraint function, we obtain the overall system performance evaluation function for the transmission efficiency and admittance characteristics of the rotating ultrasonic processing system under variable load conditions:

[0046]

[0047] In the formula C pp C sp ,L ps ,L ss The parameters are the control variables of this evaluation function, H, which ensures that the system circuit and the ultrasonic transducer operate at similar resonant frequencies. w(i) represents the weight in the i-th comparison case, min[η(i)] represents the minimum transmission efficiency value in the i-th comparison case, k(i) represents the electromechanical coupling coefficient in the i-th load case, and Q... m (i) represents the mechanical quality factor under the i-th load condition, R0(i) represents the system resonant resistance under the i-th load condition, and f r (i) represents the resonant frequency of the system circuit under the i-th load condition, f t (i) represents the resonant frequency of the ultrasonic transducer under the i-th load condition, e is the constrained maximum value of H, and a1 and a2 are C pp The minimum and maximum values ​​of b1 and b2 are C. sp The minimum and maximum values ​​of c1 and c2 are L. ps L ss The minimum and maximum values.

[0048] S3, Intelligent Algorithm Selection:

[0049] Based on the above evaluation function and variable load parameter optimization (VLPO) model designed with the admittance circle characteristics and transmission performance of CETS, a parameter optimization model applicable to variable load conditions is constructed.

[0050] The performance of multi-objective genetic algorithms and multi-objective evolutionary algorithms will be compared and analyzed, and the optimal optimization algorithm will be selected for parameter optimization of the compensation structure. The following performance metrics were selected to better evaluate the algorithm's solution performance.

[0051] (1) Hypervolume (HV) represents the volume of the target space region enclosed by the non-dominated solution set obtained by the algorithm and the reference point. The larger the HV, the better the overall performance of the algorithm.

[0052]

[0053] Where δ represents the Lebesgue measure, a standard method for measuring the length, area, or volume of a subset in Euclidean space. s represents the number of solutions, and v... i This represents the hypervolume formed by the reference point and the i-th solution in the solution set.

[0054] (2) Spatial degree (SP) represents the minimum standard deviation of distance between each solution and other solutions. The smaller the SP value, the more uniform the solution set.

[0055]

[0056] Where, d i Let d represent the minimum distance from the i-th solution to all other solutions in P, and let d represent the distance of all d-th solutions. i The mean.

[0057] The optimization process using the NSGA-III multi-objective genetic algorithm first sets the range of independent variable values ​​and the constraint function range. Then, based on the given range of compensation network parameters, the capacitor and inductor parameters are randomly initialized into a population. The efficiency performance and multiple impedance characteristic evaluation functions obtained from different combinations of compensation network parameters are non-dominated and reference points are calculated. The solution with the highest non-dominated ranking is displayed. Iterative calculations involving crossover, mutation, and population merging are then performed to find the optimal solution set as much as possible. The final solution is determined from the optimal solution set by the crowding degree calculation. When the number of evolutionary iterations reaches the set value, the optimization process ends, and a compensation network parameter combination suitable for variable load conditions is finally obtained.

[0058] The meaning of non-dominated ranking is that, assuming two parameter combinations A and B are randomly selected, if the evaluation function value of parameter combination A is less than or equal to (at least one of them is absolutely less than) the solution of parameter combination B, then solution B is said to dominate solution A. Conversely, A dominates B. If the above conditions are not met, A and B do not dominate each other, and the non-dominated ranking levels are the same.

[0059] Based on the selected performance indicators, the reference-point-based multi-objective genetic algorithm (NSGA-III) was adopted to obtain a Pareto solution set with non-dominated sorting. The evaluation functions F in the Pareto solution set were normalized to obtain f. Then, the Euclidean distance from the four-dimensional evaluation function value of each point to other points in the Pareto solution set was calculated. The minimum Euclidean distance was selected as the spatial crowding metric d for that solution. i The solution with the highest spatial congestion is selected as the optimal solution of the Pareto solution set. The calculation expression is shown in equation (14). The value of the independent variable corresponding to the optimal solution is the best parameter combination based on the VLPO model.

[0060]

[0061] Where s is the number of solution sets and f is the normalized evaluation function.

[0062] Beneficial effects of this invention:

[0063] To address the issues of poor impedance characteristics and weak anti-interference in traditional SS-structured CETS compensation, a two-sided circuit compensation network with an LC-CL structure was designed. Compared with the traditional SS topology, the impedance of the CETS is significantly reduced under the same load, and the output power driven by the same voltage is significantly improved. This solves the problem that the transmission efficiency and output power of the traditional basic topology cannot reach a high level at the same time, which affects the system's processing performance.

[0064] To address the problem of reduced transmission performance of the compensation network caused by changes in the equivalent circuit elements of the ultrasonic transducer due to load variations during ultrasonic processing, this invention proposes a transmission performance evaluation function and evaluation functions for quality factor, admittance circle radius, and coupling coefficient from the perspectives of transmission performance and impedance characteristics. A parameter optimization model (VLPO) under variable load during ultrasonic processing is constructed, and the parameters of the compensation elements of the ultrasonic transducer under variable load are obtained using a multi-objective genetic algorithm (NSGA-III). This improves the overall transmission performance and anti-interference capability of the system under variable load conditions, while also enhancing the system's load-carrying capacity. Attached Figure Description

[0065] Figure 1 This is a structural diagram of a contactless power transmission system;

[0066] Figure 2 This is the equivalent circuit of an ultrasonic transducer;

[0067] Figure 3 For circuit compensation network;

[0068] Figure 4(a) shows the equivalent reflection circuit of the LC-CL structure, which is a simplified version of the LC-CL structure circuit.

[0069] Figure 4(b) shows a simplified equivalent reflection circuit of the LC-CL structure;

[0070] Figure 5(a) shows the admittance circle of the rotary ultrasonic machining system under unmatched time-varying load conditions;

[0071] Figure 5(b) shows the admittance circle and its characteristics of the rotating ultrasonic machining system;

[0072] Figure 6 This is a schematic diagram of the optimization process of the NSGA-III algorithm;

[0073] Figure 7 A schematic diagram of the Pareto solution set;

[0074] Figure 8(a) is a comparison of simulated data transmission efficiency under different structures and strategies;

[0075] Figure 8(b) is a comparison of the simulation data quality factors under different structures and strategies;

[0076] Figure 8(c) is a comparison of the admittance circle radii of simulation data under different structures and strategies;

[0077] Figure 8(d) is a comparison of the coupling coefficients of simulation data under different structures and strategies. Detailed Implementation

[0078] The structure of a contactless power transmission system is as follows: Figure 1 As shown, energy is provided by a signal source, and after passing through the primary compensation circuit, it is transmitted to the primary coil. The current flowing through the primary coil generates a magnetic field, and the secondary coil generates a current under the generated magnetic field. The two generate mutual inductance, and the secondary current passes through the secondary compensation network to transmit energy to the ultrasonic transducer to process the workpiece.

[0079] in, Figure 1 The branch connected in parallel with C0 can be simplified to a series connection of a purely resistive load and a reactance. When the propagation speed of the longitudinal wave is different, it will exhibit different impedance characteristics. Therefore, the reactance can be equivalent to a series connection of an inductor and a capacitor, resulting in... Figure 2 The equivalent circuit shown.

[0080] Where C0 is the static capacitance of the piezoelectric transducer, and R1, C1 and L1 represent the dynamic resistance, dynamic capacitance and dynamic inductance, respectively.

[0081] A method for optimizing compensation network parameters in an ultrasonic processing system, the specific process of which is as follows:

[0082] S1. Compensation network design and performance calculation for contactless power transmission system:

[0083] To improve the system's output power, the series compensation form on the primary side is first determined, and second-order LC compensation is performed on the secondary side. Since the reactance of each component on the secondary side is inductive after being converted to the primary side through the mutual inductance model, a capacitor is chosen as the series compensation element on the primary side. While the C-CL structure compensation can achieve a suitable compensation effect by balancing transmission efficiency and output power, it is insufficient for the high efficiency and high power requirements of the rotary ultrasonic machining system. It is also necessary to effectively improve the system's impedance characteristics and increase output power while maintaining high transmission efficiency. Therefore, based on the original compensation structure, a capacitor is connected in parallel on the primary side and an inductor is connected in series, thus forming an LC-CL compensation structure. The resulting compensation network structure is as follows: Figure 3 As shown.

[0084] The circuit structure of the primary side after the mutual inductance model is shown in Figure 4(a), and the total impedance Z0 of the two parallel branches is calculated as shown in Figure 4(b).

[0085] The decomposition yields resistance R0 and reactance X0:

[0086]

[0087] Therefore, the transmission efficiency and output power formulas of the equivalent impedance of the ultrasonic transducer in CETS are shown in Table 1.

[0088] Table 1 LC-CL Transmission Performance Formula

[0089]

[0090] S2, Optimization model for compensation element parameters:

[0091] S2.1 Transmission Performance Evaluation

[0092] The efficiency performance evaluation function, based on the transmission efficiency formula of the aforementioned compensation structure, yields the transmission efficiency η under different load conditions and different compensation network parameters. This ensures that the system achieves a high level of transmission efficiency under various load conditions, while maintaining a small efficiency deviation. To meet the requirements of high transmission efficiency and high stability under various load conditions, this invention uses an impedance analyzer to measure the equivalent circuit parameters of the ultrasonic transducer under n loads and repeats the above measurement process m times. In each measurement, p loads are randomly selected from the n loads to calculate the transmission efficiency, and the weight w of each measurement is determined using the Coefficient of Variation method.

[0093] The specific calculation process is as follows: the average system transmission efficiency A during each measurement. i and standard deviation S i for:

[0094]

[0095] η(i,j) represents the system transmission efficiency under the j-th random load during the i-th measurement.

[0096] The standard deviation describes the dispersion of the values, thus yielding the coefficient of variation V of the system transmission efficiency for each measurement. i :

[0097]

[0098] After standardization, the weight w of the i-th measurement is obtained. i :

[0099]

[0100] The efficiency evaluation function F1, which determines the optimal transmission efficiency or output power performance with the smallest deviation, is:

[0101]

[0102] According to the F1 evaluation function, the greater the fluctuation of the transmission efficiency value between different load conditions in the m measurement groups, the greater the weight of that group. In order to effectively reduce the relative deviation, this invention multiplies the obtained weight by the minimum efficiency value in each measurement group, and the largest value of the calculation result is the evaluation result with the best transmission efficiency and the smallest relative deviation.

[0103] S2.2 Impedance Performance Evaluation

[0104] In addition to the transmission efficiency of CETS, the overall impedance characteristics of the system are also an important indicator for evaluating the matching effect. Therefore, this invention analyzes the admittance circle characteristics of the overall circuit of the system.

[0105] For a rotating ultrasonic machining system, under mismatched conditions, due to the presence of C0 in the equivalent circuit of the piezoelectric transducer and the inductive characteristics of the transformer coil, the admittance circle of the system exhibits a significant shift. As shown in Figure 5(a), with the increase of the ultrasonic transducer load, the dynamic resistance R1 increases, and the system resistance increases. Since the radius of the admittance circle is... When the radius of the admittance circle decreases, it will completely deviate from the real axis when the load increases to a certain level. The voltage and current of the system circuit will have a phase difference of nearly 90 degrees due to the large reactance. The active power output of the system will be close to zero, which will greatly affect the processing performance of the system.

[0106] After necessary compensation and optimization of the RUM, the overall output power of the system is effectively improved. In order to better combine the system impedance characteristics to determine the matching network parameters, this invention uses the admittance circle diagram of CETS to obtain multiple evaluation indicators, optimize the matching structure parameters, and improve the load-carrying capacity and adaptive capability of the ultrasonic processing system in actual processing. Among them, some system characteristic values ​​reflected on the admittance circle are shown in Figure 5(b), where f s For mechanical resonant frequency, f p For the parallel resonant frequency, f r For the resonant frequency, f a f1 and f2 are the anti-resonant frequencies, and f1 and f2 are the half-power points of the circuit system. Since there is still reactance in the circuit at the series resonant frequency, f s and f r The values ​​are not equal, but this can be resolved through compensation optimization. Based on some characteristic points and eigenvalues ​​of the admittance circle of the CETS circuit, several key indicators for the ultrasonic processing system are selected as evaluation functions in the selection process of compensation network parameters:

[0107] (1) Mechanical quality factor Q of the system circuit m Q is an important technical indicator for evaluating rotary ultrasonic machining systems, characterizing the energy consumed by a piezoelectric material to overcome internal friction at resonance. The lower the mechanical loss, the higher the Q value. mThe larger the value, the more its calculation expression can be represented as:

[0108]

[0109] In the formula, f1 and f2 are the half-power points of the circuit system.

[0110] The mean mechanical quality factor of the system circuit under n load conditions is used as the quality factor evaluation function F2:

[0111]

[0112] In the formula Q m (i) represents the mechanical quality factor under the i-th load condition.

[0113] (2) From the perspective of the admittance circle, a larger admittance circle radius means that the overall circuit has a lower resonant resistance, and can output higher power during rotary ultrasonic processing. To increase the admittance circle radius of the system circuit, the average admittance circle radius under n load conditions is selected as the admittance radius evaluation function F3:

[0114]

[0115] In the formula, R0(i) represents the resonant resistance of the system circuit under the i-th load condition.

[0116] (3) The electromechanical coupling coefficient is a parameter used to represent the coupling relationship between mechanical energy and electrical energy. It reflects the amount of mechanical energy that can be extracted from the total energy of the piezoelectric transducer to provide driving force, and it is also an important factor determining the bandwidth. The coupling coefficient k is related to the forward and reverse resonant frequencies f. r ,f a The calculation relationship between them is:

[0117]

[0118] Therefore, in order to improve the electromechanical coupling performance of the system, increase the system bandwidth, and enhance the system's anti-interference capability during processing, this invention uses the average electromechanical coupling coefficient of the system under n load conditions as the electromechanical coupling evaluation function F4:

[0119]

[0120] In the formula, k(i) represents the electromechanical coupling coefficient under the i-th load condition.

[0121] (4) In order to enhance the system's adaptive capability under resonant frequency shift, that is, to enable the system circuit to adaptably operate at the resonant frequency point where the oscillator is slightly shifted under different load conditions, the resonant frequency f of the CETS circuit under different load conditions is selected. rThe resonant frequency f of the ultrasonic transducer under actual conditions t The average of the differences is used as the frequency deviation constraint function H:

[0122]

[0123] In the formula f r (i) represents the resonant frequency of the rotary ultrasonic machining system circuit under the i-th load condition, f t (i) represents the resonant frequency of the ultrasonic transducer under the i-th load condition.

[0124] In the VLPO parameter optimization model decision-making process, from the perspective of the admittance circle, it is desirable for the compensation network parameters to obtain a larger system admittance circle radius and a wider bandwidth. Therefore, the compensation network parameter combination obtained through VLPO model optimization usually selects a specified C. pp A larger value is needed to achieve the desired optimization effect. Therefore, C pp The value of should be selected based on experience to choose a suitable maximum parameter, while the range of other independent variable parameters should be large enough and an appropriate step size should be chosen so that the algorithm can find the optimal parameter combination.

[0125] Finally, the above four evaluation functions and one constraint function are combined to obtain the overall system performance evaluation function for the transmission efficiency and admittance characteristics of the rotating ultrasonic processing system under variable load. The constraint range of the constraint function H should be appropriately adjusted according to the actual processing requirements. The smaller the value of H, the higher the requirement for the consistency between the CETS resonant frequency and the ultrasonic transducer resonant frequency.

[0126]

[0127] In the formula C pp C sp ,L ps ,L ss The parameters are the control variables of the evaluation function, F is a set of evaluation functions for evaluating the performance of the rotary ultrasonic machining system, and H is a constraint function that makes the system circuit and the ultrasonic transducer operate at similar resonant frequencies. w(i) represents the weight in the i-th comparison case, min[η(i)] represents the minimum transmission efficiency value in the i-th comparison case, k(i) represents the electromechanical coupling coefficient in the i-th load case, and Q... m (i) represents the mechanical quality factor under the i-th load condition, R0(i) represents the system resonant resistance under the i-th load condition, and f r (i) represents the resonant frequency of the system circuit under the i-th load condition, f t (i) represents the resonant frequency of the ultrasonic transducer under the i-th load condition, ε is the maximum constrained value of H, and a1 and a2 are C pp The minimum and maximum values ​​of b1 and b2 are C.sp The minimum and maximum values ​​of c1 and c2 are L. ps L ss The minimum and maximum values.

[0128] S3, Intelligent Algorithm Selection

[0129] Based on the above-mentioned CETS admittance circle characteristics and transmission performance, an evaluation function and variable load parameter optimization (VLPO) model are designed, and a parameter optimization model applied to variable load conditions is constructed. This invention will compare and analyze the performance of multi-objective genetic algorithms and multi-objective evolutionary algorithms, and select the optimal optimization algorithm for parameter optimization of the compensation structure. To better evaluate the algorithm's solution performance, this invention selects the following performance indicators.

[0130] (1) HV (Hypervolume) represents the volume of the target space region enclosed by the non-dominated solution set obtained by the algorithm and the reference point. The larger the HV, the better the overall performance of the algorithm.

[0131]

[0132] Where δ represents the Lebesgue measure, a standard method for measuring the length, area, or volume of a subset in Euclidean space. s represents the number of non-dominated solution sets, and v... i This represents the hypervolume formed by the reference point and the i-th solution in the solution set.

[0133] (2) SP (spacing) represents the minimum standard deviation of distance between each solution and other solutions. The smaller the SP value, the more uniform the solution set.

[0134]

[0135] Where s is the number of solutions in the solution set obtained by the algorithm, and d i Let represent the minimum distance from the i-th solution to all other solutions in P. Represents all d i The mean.

[0136] Based on the selected performance indicators, this invention employs a reference-point-based multi-objective genetic algorithm (NSGA-III) to obtain a non-dominated Pareto solution set. The objective function F of each point in the Pareto solution set is normalized to obtain f. Then, the Euclidean distance from the four-dimensional evaluation function value of each point to other points in the Pareto solution set is calculated. The minimum Euclidean distance is selected as the spatial crowding metric d of the solution. i The solution with the highest spatial congestion is selected as the optimal solution of the Pareto solution set. The calculation expression is shown in Equation (15). The optimal parameter combination is obtained based on the VLPO model.

[0137]

[0138] Where n is the number of solution sets, and f is the normalized evaluation function.

[0139] The initial population is set to 150, and the maximum population size is 10,000. In Equation 9, e = 5, a1 = 10nF, a2 = 120nF, b1 = 10nF, b2 = 100nF, c1 = 10uH, and c2 = 3000uH.

[0140] The optimization process using the NSGA-III multi-objective genetic algorithm is as follows: Figure 6 As shown, firstly, the capacitor and inductor parameters are randomly initialized according to the given range of compensation network parameters. Then, the efficiency performance and multiple impedance characteristic evaluation functions obtained by different combinations of compensation network parameters are non-dominated and reference points are calculated. The solution with the highest non-dominated ranking is displayed. Then, iterative calculations of crossover mutation and population merging operations are performed to find the optimal solution set as much as possible. The final solution is determined by the crowding degree calculation value in the optimal solution set. When the number of evolution iterations reaches the set value, the optimization process ends and the compensation network parameter combination suitable for variable load conditions is finally obtained.

[0141] The Pareto solution set obtained by the NSGA-III multi-objective genetic algorithm is as follows: Figure 7 As shown, this is a four-dimensional scatter plot. The x-axis represents the efficiency evaluation function value, the y-axis represents the quality factor evaluation function value, the z-axis represents the system admittance circle radius evaluation function value, and the color of the scatter plot represents the fourth dimension data, the evaluation function value of the effective coupling coefficient.

[0142] Based on the selected indicators, a reference-point-based multi-objective genetic algorithm (NSGA-III) is used to obtain a Pareto solution set for non-dominated sorting, as shown below. Figure 7 The objective functions F in the Pareto solution set are normalized to obtain f. Then, the Euclidean distance from the four-dimensional evaluation function value of each point to other points in the Pareto solution set is calculated. The minimum Euclidean distance is selected as the spatial crowding metric d of the solution. i The solution with the highest spatial congestion is selected as the optimal solution of the Pareto solution set. The values ​​of the independent variables corresponding to the optimal solution are obtained based on the VLPO model, which are the best parameter combinations as shown in Table 2.

[0143] Table 2 Optimal Solution of LC-CL Structural Parameters

[0144]

[0145] Comparison of data before and after LC-CL structure optimization and SS structure optimization based on VLPO model, for example Figures 8(a) to 8(d)As shown, after optimization based on the VLPO model, the LC-CL structure exhibits a slight decrease in transmission efficiency, a lower quality factor under low load but reduced volatility, and a stable performance even under high load. The admittance circle radius increases, the system resonant resistance decreases, output power is effectively increased, the coupling coefficient significantly increases and it is suitable for high load conditions, interference immunity and load-carrying capacity are greatly improved, and impedance characteristics are significantly enhanced. Based on the same VLPO model, the SS structure has slightly higher efficiency, but its impedance characteristics are insufficient. The quality factor of the CETS compensation in the SS structure is low, decreasing significantly with increasing load. It also performs poorly in terms of admittance circle radius and coupling coefficient, and large loads cause the admittance circle to completely deviate from the real axis, resulting in weak load-carrying capacity. Therefore, the LC-CL structure compensation based on the VLPO model effectively improves various impedance characteristics of the system while maintaining high efficiency, and the system's interference immunity and load-carrying capacity are effectively enhanced.

Claims

1. A method for optimizing compensation network parameters in an ultrasonic processing system, characterized in that, The specific steps include the following: S1. Compensation network design for contactless power transmission system Determine the compensation form of the primary side series compensation network, perform second-order CL compensation on the secondary side, and select a capacitor as the primary side series compensation element. Based on the original compensation structure, a capacitor element is connected in parallel on the primary side and an inductor element is connected in series to form an LC-CL compensation structure. S2, Optimization Model of Compensation Component Parameters S2.1 Transmission Performance Evaluation Measured by an impedance analyzer The equivalent circuit parameters of the ultrasonic transducer under various loads were determined, and the above measurement process was repeated. Once, in each measurement Randomly selected from various loads The transmission efficiency is calculated, and the weight of each measurement is determined using the fluctuation coefficient method. Determine the efficiency evaluation function; S2.2 Impedance Performance Evaluation Based on some characteristic points and characteristic values ​​of the admittance circle of the CETS circuit in the non-contact energy transfer system, several key indicators for the ultrasonic processing system are selected as evaluation functions and frequency deviation constraint functions in the process of selecting compensation network parameters, and the compensation network parameters are optimized. S2.3 Performance Evaluation Function By combining the evaluation function and the frequency deviation constraint function, we obtain the overall system performance evaluation function for the transmission efficiency and admittance characteristics of the rotary ultrasonic processing system under variable load conditions. S3, Intelligent Algorithm Selection Based on the evaluation function and the variable load parameter optimization model, a parameter optimization model applicable to variable load conditions is constructed: Based on the selected performance indicators, a multi-objective genetic algorithm based on reference points is adopted to obtain a Pareto solution set with non-dominated sorting. After normalizing each evaluation function in the Pareto solution set, the Euclidean distance from the four-dimensional evaluation function value of each point to other points in the Pareto solution set is calculated. The minimum Euclidean distance is selected as the measure of spatial crowding of the solution set. The solution with the highest spatial congestion is selected as the optimal solution in the Pareto solution set. The values ​​of the independent variables corresponding to the optimal solution obtained by optimizing the VLPO model based on the variable load parameters are the best parameter combinations.

2. The method for optimizing compensation network parameters of an ultrasonic processing system according to claim 1, characterized in that, The specific calculation process of the efficiency evaluation function described in S2.1 is as follows: The average system transmission efficiency during each measurement and standard deviation for: （1） in, Indicates the first During the second measurement, the first System transmission efficiency under random load; This yields the coefficient of variation of the system transmission efficiency for each measurement. : （2） After standardization, the result is the first... Weight of the second measurement : （3） An efficiency evaluation function that yields the best transmission efficiency or output power performance with the smallest deviation. for: （4） in, It is a primary-side series inductor. It is a capacitor connected in parallel on the primary side. It is a capacitor connected in parallel on the secondary side. It is a series inductor on the secondary side.

3. The method for optimizing compensation network parameters of an ultrasonic processing system according to claim 2, characterized in that, The evaluation function and frequency deviation constraint function described in S2.2 are as follows: a. Mechanical quality factor of the system circuit This characterizes the energy consumed by a piezoelectric material to overcome internal friction during resonance; the smaller the mechanical loss, the better. The larger the value, the more accurate the calculation expression becomes: （5） In the formula, and This is the half-power point of the circuit system. The resonant frequency; Will The mean mechanical quality factor of the system circuit under various load conditions is used as the quality factor evaluation function. : （6） In the formula, Indicates the first Mechanical quality factor under various load conditions; b. Selection The mean value of the admittance circle radius under various load conditions is used as the admittance radius evaluation function. : （7） In the formula, Indicates the first The resonant resistance of the system circuit under various load conditions; c. Electromechanical coupling coefficient With positive and negative resonant frequencies , The calculation relationship between them is: （8） Will The average electromechanical coupling coefficient of the system under various load conditions is used as the electromechanical coupling evaluation function. : （9） In the formula, Indicates the first Electromechanical coupling coefficient under various load conditions; d. Select The resonant frequency of the CETS circuit under various load conditions The resonant frequency of an ultrasonic transducer under actual conditions The average of the differences is used as the frequency deviation constraint function. : （10） In the formula, Indicates the first The resonant frequency of the circuit in a rotary ultrasonic machining system under various load conditions. Indicates the first The resonant frequency of the ultrasonic transducer under various load conditions.

4. The method for optimizing compensation network parameters of an ultrasonic processing system according to claim 3, characterized in that, The specific system overall performance evaluation function described in S2.3 is as follows: （11） In the formula , , , The parameters are the control variables of this evaluation function. This ensures that the system circuitry and the ultrasonic transducer operate at a frequency close to their resonant frequencies. for The maximum value of the constraint, and for The minimum and maximum values ​​are taken. and for The minimum and maximum values ​​are taken. and for , The minimum and maximum values.

5. The method for optimizing compensation network parameters of an ultrasonic processing system according to claim 4, characterized in that, Frequency deviation constraint function The constraint range is adjusted according to the actual processing conditions. The smaller the value, the higher the requirement for consistency between the CETS resonant frequency and the ultrasonic transducer resonant frequency.

6. The method for optimizing compensation network parameters of an ultrasonic processing system according to claim 1, characterized in that, The specific performance metrics described in S3 are as follows: Hypervolume HV: Represents the volume of the target space region enclosed by the non-dominated solution set obtained by the algorithm and the reference point. The larger the HV, the better the overall performance of the algorithm. （12） in, Lebesgue measure is a standard method for measuring the length, area, or volume of a subset in Euclidean space. Indicates the number of solutions. Indicates the reference point and the first solution in the solution set. The hypervolume formed by the solutions; Spatial degree (SP): represents the standard deviation of the minimum distance from each solution to all other solutions. The smaller the SP value, the more uniform the solution set. （13） in, Indicates the first A solution to the Pareto solution set The minimum distance to other solutions. Indicates all The mean.

Images