Self-tuning method for control parameters of single-phase AC input low ripple adjustable DC regulated power supply

Abstract

The present invention discloses a control parameter self-tuning method of a single-phase AC input low-ripple adjustable DC regulated power supply. The method takes the control parameter as the optimization object, takes the DC voltage ripple coefficient and steady-state control accuracy of the DC regulated power supply output as the optimization target, and establishes its multi-objective optimization fitness function; within the rated output current range, a certain current is randomly selected as the starting point, and a certain current is randomly selected as the interval to select n groups of current values ​​at equal intervals; for each group of selected output current values, the control parameter is optimized by using an improved attraction and repulsion optimization algorithm to obtain the optimal control parameter; according to the n groups of optimal control parameters, the functional relationship between each optimal control parameter and the output current is obtained; according to the functional relationship, the optimal control parameter under the output current is obtained; the DC regulated power supply is controlled by using the optimal control parameter to achieve the best operation effect of the DC regulated power supply under any load condition. The present invention has the characteristics of simple principle and good optimization effect.

Description

Technical Field

[0001] The invention belongs to the technical field of direct current regulated power supplies, and in particular relates to a control parameter self-tuning method for a single-phase alternating current input low-ripple adjustable direct current regulated power supply. Background Art

[0002] In view of the shortcomings of the current low-ripple single-phase AC input DC regulated power supply, such as complex topology and large size, the invention patent "Control method of single-phase AC input adjustable DC regulated power supply" (publication number CN117526741A) proposes a main circuit topology structure of a single-phase AC input adjustable DC regulated power supply and a corresponding control method. The DC regulated power supply has the characteristics of simple structure, small size, and arbitrarily adjustable output voltage. However, when the load of the DC regulated power supply changes, that is, when the output current of the DC regulated power supply changes, its output DC voltage ripple coefficient and steady-state control accuracy are closely related to its control parameters. Therefore, if the control parameters can be adjusted in real time according to the changes in the load of the DC regulated power supply, that is, when the output current changes, so that the output DC voltage ripple coefficient and steady-state control accuracy of the DC regulated power supply can be optimized, it will be of great significance to achieve the best operation effect of the DC regulated power supply in the whole process under variable load conditions. Summary of the invention

[0003] In order to solve the above problems existing in the prior art, the present invention provides a method for self-tuning control parameters of a single-phase AC input low-ripple adjustable DC regulated power supply.

[0004] The technical solution of the present invention to solve the above technical problems is: a method for self-tuning control parameters of a single-phase AC input low-ripple adjustable DC regulated power supply, comprising the following steps:

[0005] Step S1, for a single-phase AC input low ripple adjustable DC regulated power supply based on improved iterative learning control, taking the control parameters of the improved iterative learning control as the optimization object, taking the output DC voltage ripple coefficient and steady-state control accuracy of the DC regulated power supply as the optimization target, and establishing a mathematical model between the optimization object and the optimization target; specifically as follows:

[0006] Step S1-1, taking the inductor current i in the single-phase Buck-Boost inverter circuit L and capacitor voltage u C is the system state variable, and its state differential equation is established, specifically:

[0007] (1);

[0008] Where: i L is the inductor current, u C is the capacitor voltage, 、 i L、 u C The first derivative of DC is the DC voltage at the input side of the single-phase Buck-Boost inverter circuit, d is the duty cycle of the power switch in the single-phase Buck-Boost inverter circuit, L and C are the inductance and capacitance in the single-phase Buck-Boost inverter circuit respectively, R req is the equivalent load resistance.

[0009] Step S1-2, the improved iterative learning control method is used to control the single-phase Buck-Boost inverter circuit. According to the basic principle of the control method, the duty cycle d(s) of the corresponding power switch tube in the inverter circuit is obtained as follows:

[0010] (2);

[0011] Where: E(s) is the image function of the deviation between the actual output DC voltage of the DC regulated power supply and its reference output voltage, k Ip , k Ii , k F are the gain coefficient, integral coefficient and filter gain coefficient of the improved iterative learning control respectively, ω is the angular frequency of the output AC voltage of the single-phase Buck-Boost inverter circuit, s is the complex variable of Laplace transform, τ and α are constant coefficients, and e is a natural constant.

[0012] Step S1-3, according to equation (1) and equation (2), the capacitor voltage u in the single-phase Buck-Boost inverter circuit is obtained. C (s), specifically:

[0013] (3);

[0014] Step S1-4, according to formula (3) and the structural characteristics of the single-phase Buck-Boost inverter circuit, the output voltage u of the single-phase Buck-Boost inverter circuit is obtained. PN (s) is:

[0015] (4);

[0016] Where: U OD is the DC bias component of the capacitor voltage.

[0017] Step S1-5, according to formula (4), combined with the transformation processing of the single-phase bridge uncontrolled rectifier circuit and the π-type RLC combined filter circuit, and then through the Laplace inverse transformation, the analytical expression of the DC voltage output by the DC regulated power supply is obtained, which is specifically:

[0018] (5);

[0019] Where: t is the system running time, L f1 , L f2 , C f1 , C f2 , R f1 are the first and second filter inductors, the first and second filter capacitors and the damping resistor of the π-type RLC combined filter circuit, A 1 , A 2 , A 3 All are constant coefficients, among which:

[0020] (6);

[0021] (7);

[0022] Step S1-6, according to formula (5), the functional relationship between the DC voltage ripple coefficient r and the steady-state control accuracy η of the DC regulated power supply output is obtained as follows:

[0023] (8);

[0024] Where: E 0 It is the initial value of the deviation between the actual output DC voltage and the reference output voltage;

[0025] (9);

[0026] Where: Output DC voltage ripple coefficient r = peak value of AC component in output DC voltage / average value of output DC voltage; Output DC voltage steady-state control accuracy η = (output DC reference voltage - output DC average voltage) / output DC reference voltage × 100%, U dc_ref Outputting a DC reference voltage for a DC regulated power supply;

[0027] Step S2, based on the mathematical model established in step S1, a multi-objective optimization fitness function is established using a linear programming method;

[0028] Step S3, randomly selecting a certain current within the rated output current range of the DC regulated power supply as a starting point, and randomly selecting n groups of current values ​​at equal intervals from a certain current;

[0029] Step S4, for each selected group of output current values, optimizing the control parameters thereof by using an improved attraction-repulsion optimization algorithm, and obtaining the optimal control parameters corresponding to the output current;

[0030] Step S5, according to the obtained n groups of optimal control parameters and their corresponding output current values, a numerical fitting method is used to obtain the functional relationship between each optimal control parameter and its output current;

[0031] Step S6, obtaining the optimal control parameters of the DC regulated power supply under the output current according to the obtained functional relationship of each optimal control parameter and the actual output current value of the DC regulated power supply;

[0032] Step S7, controlling the DC regulated power supply according to the obtained optimal control parameters.

[0033] Compared with the prior art, the present invention has the following beneficial effects:

[0034] (1) The present invention improves the traditional attraction-repulsion optimization algorithm by using the Fuch chaotic mapping function to improve its population initialization formula in its population initialization stage, and then using a perturbation strategy based on fitness classification and a directional perturbation strategy based on the population center in the population update stage, thereby effectively improving the optimization efficiency and optimization accuracy of the algorithm.

[0035] (2) The present invention adopts an improved attraction-repulsion optimization algorithm to optimize the control parameters of the single-phase AC input low ripple adjustable DC regulated power supply using improved iterative learning control. Compared with the traditional attraction-repulsion optimization algorithm, it effectively reduces the output DC voltage ripple coefficient and steady-state error of the DC regulated power supply, and achieves significant optimization effect.

[0036] (3) According to the functional relationship between the optimal control parameters of the single-phase AC input low-ripple adjustable DC regulated power supply based on improved iterative learning control and its output current, the optimal control parameters of the DC regulated power supply under any output current can be obtained. By controlling the DC regulated power supply according to the obtained optimal control parameters, the DC regulated power supply can achieve the optimal values ​​of its output DC voltage ripple coefficient and steady-state control accuracy under any load condition. The present invention is easy to implement and has a very good optimization effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 This is a topological diagram of the main circuit of the single-phase AC input low ripple adjustable DC regulated power supply in the present invention.

[0038] Figure 2 The present invention is a flow chart of the method for self-tuning control parameters of a single-phase AC input low ripple adjustable DC regulated power supply.

[0039] Figure 3 This is a flow chart for optimizing control parameters of a DC regulated power supply based on an improved attraction-repulsion optimization algorithm in the present invention.

[0040] Figure 4 is the optimal control parameter k of the single-phase AC input low ripple adjustable DC regulated power supply in the present invention. Ip Fitting curve graph.

[0041] Figure 5 is the optimal control parameter k of the single-phase AC input low ripple adjustable DC regulated power supply in the present invention. Ii Fitting curve graph.

[0042] Figure 6 is the optimal control parameter k of the single-phase AC input low ripple adjustable DC regulated power supply in the present invention. F Fitting curve graph. DETAILED DESCRIPTION

[0043] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0044] See also Figure 1 , which is a topological structure diagram of the main circuit of the single-phase AC input low-ripple adjustable DC regulated power supply provided by the present invention. The topological structure includes four parts: a single-phase PWM rectifier circuit, a single-phase Buck-Boost inverter circuit, a single-phase bridge uncontrolled rectifier circuit and a π-type RLC combined filter circuit. Its basic working principle is: the single-phase PWM rectifier circuit rectifies the single-phase AC input voltage into a PWM-modulated DC voltage, and the single-phase Buck-Boost inverter circuit inverts the PWM-modulated DC voltage output by the single-phase PWM rectifier circuit into a high-quality sinusoidal AC voltage with adjustable amplitude and frequency. The sinusoidal AC voltage is then rectified by a single-phase bridge uncontrolled rectifier circuit and filtered by a π-type RLC combined filter circuit, and finally outputs a low-ripple adjustable DC voltage. Among them, the single-phase Buck-Boost inverter circuit is composed of two groups of Buck-Boost DC-DC conversion circuits with the same structure connected in parallel in a phase complementary manner. The single-phase Buck-Boost inverter circuit is controlled by an improved iterative learning control method and can directly output a high-quality sinusoidal AC voltage with arbitrarily adjustable amplitude and frequency.

[0045] See also Figure 2 and Figure 3 , respectively, are a flow chart of a single-phase AC input low-ripple adjustable DC regulated power supply control parameter self-tuning method provided by the present invention and a flow chart of a DC regulated power supply control parameter optimization based on an improved attraction-repulsion optimization algorithm, specifically comprising the following steps:

[0046] Step S1, for a single-phase AC input low ripple adjustable DC regulated power supply based on improved iterative learning control, taking the control parameters of the improved iterative learning control as the optimization object, taking the output DC voltage ripple coefficient and steady-state control accuracy of the DC regulated power supply as the optimization target, and establishing a mathematical model between the optimization object and the optimization target, the specific steps are as follows:

[0047] Step S1-1, taking the inductor current i in the single-phase Buck-Boost inverter circuit L and capacitor voltage u C is the system state variable, and its state differential equation is established, specifically:

[0048] (1);

[0049] Where: i L is the inductor current, u C is the capacitor voltage, 、 i L、 u C The first derivative of DC is the DC voltage at the input side of the single-phase Buck-Boost inverter circuit, d is the duty cycle of the power switch in the single-phase Buck-Boost inverter circuit, L and C are the inductance and capacitance in the single-phase Buck-Boost inverter circuit respectively, R req is the equivalent load resistance.

[0050] Step S1-2, the improved iterative learning control method is used to control the single-phase Buck-Boost inverter circuit. According to the basic principle of the control method, the duty cycle d(s) of the corresponding power switch tube in the inverter circuit is obtained as follows:

[0051] (2);

[0052] Where: E(s) is the image function of the deviation between the actual output DC voltage of the DC regulated power supply and its reference output voltage, k Ip , k Ii , k F are the gain coefficient, integral coefficient and filter gain coefficient of the improved iterative learning control respectively, ω is the angular frequency of the output AC voltage of the single-phase Buck-Boost inverter circuit, s is the complex variable of Laplace transform, τ and α are constant coefficients, and e is a natural constant.

[0053] Step S1-3, according to equation (1) and equation (2), the capacitor voltage u in the single-phase Buck-Boost inverter circuit is obtained. C (s), specifically:

[0054] (3);

[0055] Step S1-4, according to formula (3) and the structural characteristics of the single-phase Buck-Boost inverter circuit, the output voltage u of the single-phase Buck-Boost inverter circuit is obtained. PN (s) is:

[0056] (4)

[0057] Where: U OD is the DC bias component of the capacitor voltage.

[0058] Step S1-5, according to formula (4), combined with the transformation processing of the single-phase bridge uncontrolled rectifier circuit and the π-type RLC combined filter circuit, and then through the Laplace inverse transformation, the analytical expression of the DC voltage output by the DC regulated power supply is obtained, which is specifically:

[0059] (5);

[0060] Where: t is the system running time, L f1 , L f2 , C f1 , C f2 , R f1 are the first and second filter inductors, the first and second filter capacitors and the damping resistor of the π-type RLC combined filter circuit, A 1 , A 2 , A 3 All are constant coefficients, among which:

[0061] (6);

[0062] (7);

[0063] Step S1-6, according to formula (5), the functional relationship between the DC voltage ripple coefficient r and the steady-state control accuracy η of the DC regulated power supply output is obtained as follows:

[0064] (8);

[0066] Where: E 0 It is the initial value of the deviation between the actual output DC voltage and the reference output voltage;

[0067] (9);

[0068] Where: Output DC voltage ripple coefficient r = peak value of AC component in output DC voltage / average value of output DC voltage; Output DC voltage steady-state control accuracy η = (output DC reference voltage - output DC average voltage) / output DC reference voltage × 100%, U dc_ref Outputs DC reference voltage for DC regulated power supply.

[0069] Step S2, based on the mathematical model established in step S1, a linear programming method is used to establish the fitness function of its multi-objective optimization:

[0070] (10);

[0071] Where: f(k Ip ,k Ii ,k F ) is the multi-objective optimization fitness function for DC regulated power supply control parameter optimization, k 1 is the weight coefficient of the output DC voltage ripple coefficient, k 2 is the weight coefficient of the output DC voltage steady-state control accuracy.

[0072] Step S3, randomly selecting a certain current within the rated output current range of the DC regulated power supply as a starting point, and randomly selecting n groups of current values ​​at equal intervals from a certain current;

[0073] Step S4, for each selected set of output current values, the control parameters are optimized using an improved attraction and repulsion optimization algorithm to obtain the optimal control parameters corresponding to the output current, as follows:

[0074] The improved attraction-repulsion optimization algorithm is to make the following improvements to the traditional attraction-repulsion optimization algorithm:

[0075] 1) In the population initialization stage, the Fuch chaos mapping function is used to improve the population initialization formula, so that the individuals of the initial population are more evenly distributed, the omission of the search area is eliminated, and the local optimal solution is avoided. The improved population initialization formula is:

[0076] (11);

[0077] Where: x i is the i-th individual in the population, m is the current iteration number, x(m+1) is the Fuch chaos mapping function, U b , L b are the upper and lower boundaries of the population search space respectively.

[0078] 2) In the population update stage, the perturbation strategy based on fitness classification and the directional perturbation strategy based on population center are used for improvement, as follows:

[0079] 2-1) Before the number of iterations reaches 30% of the maximum number of iterations or when the standard deviation of the fitness values ​​of all individuals in the population is greater than 0.01, a perturbation strategy based on fitness grading is introduced, that is, different intensities of perturbations are applied to each individual according to the fitness value grading, thereby effectively improving the global search ability of the population. Specifically:

[0080] First, the individuals in the population are graded according to their fitness values, that is, they are sorted from small to large according to the fitness values ​​of each individual in the population. The first 30% of the individuals in the sequence are determined to be of high quality, the last 30% of the individuals in the sequence are of low quality, and the rest are of medium quality.

[0081] Then, according to the above classification of population individuals, different intensities of disturbances are applied to them, specifically:

[0082] (12);

[0083] Where: x i (m-1) is the population individual of the m-1th iteration, β 1 , β 2 , β 3 are the level disturbance intensities of individuals in high-quality, medium-quality, and low-quality populations, respectively, where β 1 is a constant less than 1, β 3 is a constant greater than 1, β 2 ≈1, N(0,1) represents a random number that follows a normal distribution, and rand is a random number in the range [0,1].

[0084] 2-2) When the standard deviation of the fitness values ​​of all individuals in the population is less than 0.001, a directional perturbation strategy based on the population center is introduced to guide the individuals in the population to search in the direction of the global optimal solution, avoid falling into the local optimal solution, and effectively improve the convergence speed and optimization accuracy of the optimization. Specifically:

[0085] Calculate the deviation between the current individual and the population center, specifically:

[0086] (13);

[0087] Where: , is the population center value, x i is the population individual, and N is the population size.

[0088] According to formula (13), the disturbance is applied in the population search direction, specifically:

[0089] (14);

[0090] Where: γ is the disturbance coefficient, δ is the random disturbance factor, as follows:

[0091] (15);

[0092] Where: γ 0 is the initial value of the perturbation coefficient, δ 0 is the initial value of the random perturbation factor, and M is the maximum number of iterations.

[0093] The control parameters of the DC regulated power supply are optimized by using an improved attraction-repulsion optimization algorithm, which specifically includes the following steps:

[0094] S4-1, set the number of individuals in the initial population N, the maximum number of iterations M, the population dimension dim, and the upper and lower boundaries of the population search U b and L b , take the initial value of the perturbation coefficient γ 0 , initial value of random perturbation factor δ 0 Both are 0.5;

[0095] S4-2, randomly initialize the population X;

[0096] S4-3, calculate the individual x in the initial population X according to formula (10) i The fitness value f(x i ), take the minimum fitness value f b The corresponding individual is the current optimal individual x b , and f b and x b Save to register F m and X m middle;

[0097] S4-4, update the population according to the improved attraction and repulsion optimization algorithm to generate the next generation of new population X new ;

[0098] S4-5, calculate the x of each individual in the new population i new The fitness value f(x i new ), determine its minimum fitness value f b new Is it less than the currently saved minimum fitness value f? b , if so, then the currently saved minimum fitness value f b and the corresponding optimal individual x b Respectively b new and x b new Replace; otherwise, the currently saved minimum fitness value and the corresponding optimal individual remain unchanged;

[0099] S4-6, determine whether the maximum number of iterations M has been reached; if so, execute step S4-7; otherwise, increase the number of iterations by 1 and return to step S4-4;

[0100] S4-7, output the optimal individual value x b, which is to improve the optimal value of each control parameter of iterative learning control.

[0101] Step S5, according to the obtained n groups of optimal control parameters and their corresponding output current values, a numerical fitting method is used to obtain a functional relationship between each optimal control parameter and its output current, which is specifically as follows:

[0102] Optimal control parameter k Ip With its output current I o The functional relationship between Ip (I o )for:

[0103] ;

[0104] Optimal control parameter k Ii With its output current I o The functional relationship between Ii (I o )for:

[0105]

[0106] Optimal control parameter k F With its output current I o The functional relationship between F (I o )for:

[0107]

[0108] Where: a 0 、a 1 、a 2 They are the optimal control parameter function k Ip (I o ), b 0 、b 1 、b 2 、b 3 、b 4 They are the optimal control parameter function k Ii (I o ), c 0 、c 1 、c 2 、c 3 、c 4 They are the optimal control parameter function k F (I o ) in the coefficients.

[0109] Step S6, according to the obtained functional relationship of each optimal control parameter and the actual output current value of the DC regulated power supply, the optimal control parameter of the power supply under the output current can be obtained;

[0110] Step S7, controlling the DC regulated power supply according to the obtained optimal control parameters, can achieve the operating effect that the output DC voltage ripple coefficient and steady-state control accuracy of the DC regulated power supply can reach the optimal values ​​under any load condition.

[0111] In this embodiment, in order to verify the effect of the self-tuning method for control parameters of the single-phase AC input low ripple adjustable DC regulated power supply provided by the present invention, the main technical parameters of the DC regulated power supply are shown in Table 1, and the relevant parameters of the improved attraction and repulsion optimization algorithm adopted are shown in Table 2.

[0112] Table 1 Main technical parameters of single-phase AC input low ripple adjustable DC regulated power supply

[0113]

[0114] Table 2 Parameters of the improved attraction-repulsion optimization algorithm

[0115]

[0116] Six groups of output current values ​​are randomly selected at equal intervals within the rated output current range of the DC regulated power supply, such as 1A, 2A, 3A, 4A, 5A and 6A respectively; for each group of output current values ​​selected, according to the parameters shown in Table 2, the improved attraction-repulsion optimization algorithm is used to optimize the control parameters of the improved iterative learning control adopted by the DC regulated power supply, and the corresponding optimal control parameters are shown in Table 3.

[0117] Table 3 Optimal control parameters under different output currents

[0118]

[0119] According to the optimal control parameters and their corresponding output current values ​​obtained in Table 3, the functional relationship between the optimal control parameters and their output current is obtained by numerical fitting method, which is as follows:

[0120] Optimal control parameter k Ip With its output current I o The functional relationship between Ip (I o )for:

[0121]

[0122] Optimal control parameter k Ii With its output current I o The functional relationship between Ii (I o )for:

[0123]

[0124] Optimal control parameter k F With its output current I o The functional relationship between F (I o )for:

[0125]

[0126] According to the data obtained in Table 3 and the functional relationships shown in equations (19)-(21), the corresponding fitting curves are obtained as follows: Figure 4~Figure 6 shown.

[0127] In order to verify the effect of the functional relationship between the optimal control parameters of the DC regulated power supply and its output current, if the output current values ​​of the DC regulated power supply are 1.5A and 3.5A respectively, for these two groups of output current values, the functional relationship shown in equations (19)-(21) is used for calculation and the improved attraction-repulsion optimization algorithm is directly used for optimization, and the corresponding optimal control parameters under the two methods are obtained. Then, the output DC voltage ripple coefficient and steady-state control accuracy of the DC regulated power supply are obtained according to the obtained optimal control parameters, as shown in Table 4.

[0128] Table 4 Output DC voltage ripple coefficient and steady-state control accuracy corresponding to the two methods

[0129]

[0130] It can be seen from Table 4 that the output DC voltage ripple coefficient and steady-state control accuracy corresponding to the above two methods are basically consistent, and their maximum relative errors are 2.0% and 1.1% respectively, which further verifies the effectiveness of the optimal control parameter function relationship provided by the present invention.

[0131] At the same time, in order to further verify the effect of the control parameter self-tuning method of the single-phase AC input low ripple adjustable DC regulated power supply provided by the present invention, the present invention is compared with the traditional control method using fixed control parameters (referred to as the traditional method) for analysis, wherein the traditional method uses the optimal control parameters corresponding to the rated operating point for control. For example, if the output current values ​​are 1.6A and 2.5A respectively, the present invention and the traditional method are used for control for these two groups of output current values, respectively, and the output DC voltage ripple coefficient and steady-state control accuracy corresponding to the two methods are obtained, as shown in Table 5.

[0132] Table 5 Output DC voltage ripple coefficient and steady-state control accuracy corresponding to the two methods

[0133]

[0134] It can be seen from Table 5 that when the output current of the power supply is 1.6A and 2.5A respectively, the output DC voltage ripple coefficient obtained by the present invention is reduced by 26.7% and 22.3% respectively compared with the traditional method, and the steady-state control accuracy is reduced by 20% and 18.2% respectively, which further verifies the effectiveness of the present invention.

Claims

1. A method for self-tuning control parameters of a single-phase AC input low ripple adjustable DC regulated power supply, characterized in that: The following steps are involved: Step S1, for a single-phase AC input low ripple adjustable DC regulated power supply based on improved iterative learning control, taking the control parameters of the improved iterative learning control as the optimization object, taking the output DC voltage ripple coefficient and steady-state control accuracy of the DC regulated power supply as the optimization target, and establishing a mathematical model between the optimization object and the optimization target; specifically as follows: Step S1-1, taking the inductor current i in the single-phase Buck-Boost inverter circuit L and capacitor voltage u C is the system state variable, and its state differential equation is established, specifically: (1)； Where: i L is the inductor current, u C is the capacitor voltage, 、 i L、 u C The first derivative of DC is the DC voltage at the input side of the single-phase Buck-Boost inverter circuit, d is the duty cycle of the power switch in the single-phase Buck-Boost inverter circuit, L and C are the inductance and capacitance in the single-phase Buck-Boost inverter circuit respectively, R req is the equivalent load resistance; Step S1-2, the improved iterative learning control method is used to control the single-phase Buck-Boost inverter circuit. According to the basic principle of the control method, the duty cycle d(s) of the corresponding power switch tube in the inverter circuit is obtained as follows: (2)； Where: E(s) is the image function of the deviation between the actual output DC voltage of the DC regulated power supply and its reference output voltage, k Ip , k Ii , k F are the gain coefficient, integral coefficient and filter gain coefficient of the improved iterative learning control, ω is the angular frequency of the output AC voltage of the single-phase Buck-Boost inverter circuit, s is the complex variable of Laplace transform, τ and α are constant coefficients, and e is a natural constant; Step S1-3, according to equation (1) and equation (2), the capacitor voltage u in the single-phase Buck-Boost inverter circuit is obtained. C (s), specifically: (3)； Step S1-4, according to formula (3) and the structural characteristics of the single-phase Buck-Boost inverter circuit, the output voltage u of the single-phase Buck-Boost inverter circuit is obtained. PN (s) is: (4)； Where: U OD is the DC bias component of the capacitor voltage; Step S1-5, according to formula (4), combined with the transformation processing of the single-phase bridge uncontrolled rectifier circuit and the π-type RLC combined filter circuit, and then through the Laplace inverse transformation, the analytical expression of the DC voltage output by the DC regulated power supply is obtained, which is specifically: (5)； Where: t is the system running time, L f1 , L f2 , C f1 , C f2 , R f1 They are the first and second filter inductors, the first and second filter capacitors and the damping resistor of the π-type RLC combined filter circuit, A1, A2 and A3 are all constant coefficients, where: (6)； (7)； Step S1-6, according to formula (5), the functional relationship between the DC voltage ripple coefficient r and the steady-state control accuracy η of the DC regulated power supply output is obtained as follows: (8)； Where: E0 is the initial value of the deviation between the actual output DC voltage and the reference output voltage; (9)； Where: Output DC voltage ripple coefficient r = peak value of AC component in output DC voltage / average value of output DC voltage; Output DC voltage steady-state control accuracy η = (output DC reference voltage - output DC average voltage) / output DC reference voltage × 100%, U dc_ref Outputting a DC reference voltage for a DC regulated power supply; Step S2, based on the mathematical model established in step S1, a multi-objective optimization fitness function is established using a linear programming method; Step S3, randomly selecting a certain current within the rated output current range of the DC regulated power supply as a starting point, and randomly selecting n groups of current values ​​at equal intervals from a certain current; Step S4, for each selected group of output current values, optimizing the control parameters thereof by using an improved attraction-repulsion optimization algorithm, and obtaining the optimal control parameters corresponding to the output current; Step S5, according to the obtained n groups of optimal control parameters and their corresponding output current values, a numerical fitting method is used to obtain the functional relationship between each optimal control parameter and its output current; Step S6, obtaining the optimal control parameters of the DC regulated power supply under the output current according to the obtained functional relationship of each optimal control parameter and the actual output current value of the DC regulated power supply; Step S7, controlling the DC regulated power supply according to the obtained optimal control parameters.

2. A method for self-tuning control parameters of a single-phase AC input low ripple adjustable DC regulated power supply according to claim 1, characterized in that: The single-phase AC input adjustable DC regulated power supply comprises a single-phase PWM rectifier circuit, a single-phase Buck-Boost inverter circuit, a single-phase bridge uncontrollable rectifier circuit and a π-type RLC combined filter circuit.

3. The method for self-tuning control parameters of a single-phase AC input low ripple adjustable DC regulated power supply according to claim 1, characterized in that: The multi-objective optimization fitness function in step S2 is: (10)； Where: f(k Ip ,k Ii ,k F ) is the multi-objective optimization fitness function for optimizing the control parameters of the DC regulated power supply, k1 is the weight coefficient of the output DC voltage ripple coefficient, and k2 is the weight coefficient of the output DC voltage steady-state control accuracy.

4. The method for self-tuning control parameters of a single-phase AC input low ripple adjustable DC regulated power supply according to claim 1, characterized in that: The improved attraction-repulsion optimization algorithm in step S4 is to make the following improvements to the traditional attraction-repulsion optimization algorithm: 1) In the population initialization stage, the Fuch chaos mapping function is used to improve the population initialization formula. The improved population initialization formula is: (11)； Where: x i is the i-th individual in the population, m is the current iteration number, x(m+1) is the Fuch chaos mapping function, U b , L b are the upper and lower boundaries of the population search space respectively; 2) In the population update stage, the perturbation strategy based on fitness classification and the directional perturbation strategy based on population center are used for improvement, as follows: 2-1) Before the number of iterations reaches 30% of the maximum number of iterations or when the standard deviation of the fitness values ​​of all individuals in the population is greater than 0.01, a perturbation strategy based on fitness classification is introduced; specifically: The individuals in the population are graded according to their fitness values, that is, they are sorted from small to large according to their fitness values. The first 30% of the individuals in the sequence are determined to be of high quality, the last 30% of the individuals in the sequence are of low quality, and the rest are of medium quality. Then, according to the above classification of population individuals, different intensities of disturbances are applied to them, specifically: (12)； Where: x i (m-1) is the population individual of the m-1th iteration, β1, β2, and β3 are the disturbance intensities of high-quality, medium-quality, and low-quality population individuals, respectively, where β1 is a constant less than 1, β3 is a constant greater than 1, β2≈1, N(0,1) represents a random number that obeys a normal distribution, and rand is a random number in the range [0,1]; 2-2) When the standard deviation of the fitness values ​​of all individuals in the population is less than 0.001, a directional perturbation strategy based on the population center is introduced to guide the individuals in the population to search in the direction of the global optimal solution to avoid falling into the local optimal solution; specifically: Calculate the deviation between the current individual and the population center, specifically: (13)； Where: , is the population center value, x i is the population individual, N is the population size; According to formula (13), the disturbance is applied in the population search direction, specifically: (14)； Where: γ is the disturbance coefficient, δ is the random disturbance factor, as follows: (15)； Where: γ0 is the initial value of the perturbation coefficient, δ0 is the initial value of the random perturbation factor, and M is the maximum number of iterations.

5. The method for self-tuning control parameters of a single-phase AC input low ripple adjustable DC regulated power supply according to claim 1, characterized in that: In step S4, the control parameters of the DC regulated power supply are optimized by using an improved attraction and repulsion optimization algorithm, which specifically includes the following steps: S4-1, set the number of individuals in the initial population N, the maximum number of iterations M, the population dimension dim, and the upper and lower boundaries of the population search U b and L b , the initial value of the disturbance coefficient γ0 and the initial value of the random disturbance factor δ0 are both 0.5; S4-2, randomly initialize the population X; S4-3, calculate the individual x in the initial population X according to formula (10) i The fitness value f(x i ), i=1,···,N, and take the one with the smallest fitness value f b The corresponding individual is the current optimal individual x b , and f b and x b Save to register F m and X m middle; S4-4, update the population according to the improved attraction and repulsion optimization algorithm to generate the next generation of new population X new ; S4-5, calculate the x of each individual in the new population i new The fitness value f(x i new );Determine its minimum fitness value f b new Is it less than the currently saved minimum fitness value f? b , if so, then the currently saved minimum fitness value f b and the corresponding optimal individual x b Respectively b new and x b new Replace; otherwise, the currently saved minimum fitness value and the corresponding optimal individual remain unchanged; S4-6, determine whether the maximum number of iterations M has been reached; if so, execute step S4-7; otherwise, increase the number of iterations by 1 and return to step S4-4; S4-7, output the optimal individual value x b , which is to improve the optimal value of each control parameter of iterative learning control.

6. The method for self-tuning control parameters of a single-phase AC input low ripple adjustable DC regulated power supply according to claim 1, characterized in that: The functional relationship between the optimal control parameters of the DC regulated power supply and its output current obtained in step S5 is as follows: Optimal control parameter k Ip With its output current I o The functional relationship between Ip (I o )for: ； Optimal control parameter k Ii With its output current I o The functional relationship between Ii (I o )for: ； Optimal control parameter k F With its output current I o The functional relationship between F (I o )for: ； Where: a0, a1, a2 are the optimal control parameter function k Ip (I o ), b0, b1, b2, b3, b4 are the optimal control parameter function k Ii (I o ), c0, c1, c2, c3, c4 are the optimal control parameter function k F (I o ) in the coefficients.

Images