Model-free adaptive oscillation suppression method and equipment for wind power flexible direct current grid-connected system with data disturbance observer, computer equipment and readable storage medium

Abstract

The invention discloses a model-free adaptive oscillation suppression method and equipment for a wind power flexible direct current grid-connected system with a data disturbance observer, computer equipment and a readable storage medium. According to the method, a model-free adaptive oscillation suppression strategy and a wind power flexible direct current grid-connected system are integrated, a unique oscillation suppression structure is constructed, uncertain factors in the system can be effectively coped with by means of a data disturbance compensation mechanism, and the stability of the system under complex working conditions is improved. By obtaining the optimized strategy position and parameters, the operation of the wind power plant can be more accurately controlled, the reliability and safety of wind power integration are enhanced, the problems of fan shutdown, equipment damage and the like caused by oscillation are avoided, and stable operation of a power system is guaranteed.

Description

Technical Field

[0001] The present invention relates to a wide range of areas, and in particular to a method, equipment, computer equipment and readable storage medium for model-free adaptive oscillation suppression of a wind power flexible direct current grid-connected system with a data disturbance observer. Background Art

[0002] As the global energy mix accelerates toward clean energy, wind power, thanks to its renewable and pollution-free nature, has become a major force in the new energy sector. The accompanying flexible direct current (HVDC) transmission technology, with its significant advantages such as flexible control, large transmission capacity, and low transmission losses, is becoming an increasingly critical method for delivering wind power.

[0003] The control flexibility of HVDC Flexible technology lies in its ability to rapidly and precisely regulate active and reactive power through advanced power electronics and control strategies. When wind power output fluctuates, the HVDC Flexible system can adjust transmission power in real time to ensure stable grid operation. For example, when wind speed fluctuations cause wind turbine output power to fluctuate, the HVDC Flexible system can respond within milliseconds, preventing power fluctuations from impacting the grid. Its large transmission capacity enables it to meet the long-distance power transmission needs of large-scale wind farms. For example, large wind farms in western my country can transmit thousands of megawatts of wind power to load centers thousands of kilometers away in central and eastern China through HVDC Flexible lines, significantly increasing wind power absorption capacity. Furthermore, compared to traditional AC transmission, HVDC Flexible transmission can reduce losses by 20%-30% during long-distance transmission, improving energy efficiency and reducing transmission costs.

[0004] However, with the continuous expansion of wind power grid-connected via flexible DC transmission, broadband oscillations have become a frequent problem, hindering the efficient and stable utilization of wind power. Broadband oscillations are power and voltage oscillations with a wide frequency range (typically ranging from a few hertz to several hundred hertz). The causes of these oscillations are complex. On the one hand, there is an interaction between the wind turbine's power electronic converter and the control circuit of the flexible DC transmission system. When the control parameters of the two are mismatched, oscillations can easily occur. On the other hand, the coupling between the mechanical dynamic characteristics of the wind turbine and the dynamic characteristics of the electrical system can also lead to broadband oscillations. Once these oscillations occur, they can easily cause wind turbine shutdowns, resulting in interruptions in wind power output. Frequent oscillations can also cause mechanical and electrical stress on transformers, reactors, and other equipment in the system, accelerating equipment aging and even causing damage. More seriously, if the oscillations cannot be effectively suppressed, they can trigger cascading failures such as disconnection of the wind power grid-connected via flexible DC transmission system, posing a serious threat to the safe and stable operation of the entire power system.

[0005] Currently, most methods for suppressing oscillations in wind power grid-connected flexible direct current (DC) systems employ model-driven control methods. For example, some studies establish detailed mathematical models of the system and employ strategies such as state feedback control and robust control to suppress oscillations. However, the structure of a wind power grid-connected flexible direct current (DC) system is complex, encompassing multiple links such as wind turbines, converter stations, and transmission lines, and the system exhibits strong nonlinear characteristics. This complexity and nonlinearity inevitably lead to unmodeled dynamic links and other uncertainties in model-driven oscillation suppression methods. During actual operation, system parameters may change, such as changes in wind turbine mechanical parameters due to wear or changes in the system's operating conditions. These changes can reduce the accuracy of the model and, in turn, affect the effectiveness of oscillation suppression.

[0006] It can be seen that the current model-driven method for suppressing oscillations in wind power systems connected to the flexible direct current grid needs further improvement. Summary of the Invention

[0007] To avoid and overcome the technical problems existing in the prior art, the present invention provides a model-free adaptive oscillation suppression method, equipment, computer device, and readable storage medium for a wind power grid-connected flexible direct current system with a data disturbance observer. The method of the present invention enables more precise control of wind farm operation.

[0008] To achieve the above object, the present invention provides the following technical solutions:

[0009] A model-free adaptive oscillation suppression method for a wind power flexible direct current grid-connected system with a data disturbance observer comprises the following steps:

[0010] A1. Based on the model-free adaptive oscillation suppression strategy and the wind power grid-connected flexible DC system, a model-free adaptive oscillation suppression structure for the wind power grid-connected flexible DC system is established;

[0011] A2. Compensate for data disturbances in the wind power grid-connected flexible direct current system by calculating the control rate, total disturbance estimate, and pseudo-partial derivative including disturbance compensation.

[0012] A3. Obtain the position of the model-free adaptive oscillation suppression strategy in the wind power flexible direct current grid-connected system after data disturbance compensation, and perform corresponding parameter optimization;

[0013] A4. The operation of wind farms is controlled by the flexible direct current grid-connected system with optimized wind power parameters.

[0014] As a further solution of the present invention, the wind power grid-connected system via flexible direct current is represented as follows:

[0015] y(k+1)=f(y(k),...,y(kn y ),u(k),...,u(kn u))+d(k);

[0016] Where u(k) and y(k) are the input and output of the system at time k, respectively; y(k+1) is the output of the system at time k+1; u(k–n u ) and y(k–n y ) are the system in k–n u The input and k–n at the moment y Output at time; n u and n y are the input order and output order of the system respectively; f(·) is the nonlinear function that characterizes the characteristics of the system; d(k) is the bounded unknown disturbance received by the system at time k.

[0017] As a further solution of the present invention: the calculation process of the control rate is as follows:

[0018] First, the system representation is converted into an iterative representation that can be iteratively calculated. The specific representation is as follows:

[0019] y(k+1)=y(k)+φ c (k)Δu(k)+Δd(k);

[0020] Where, φ c (k) is the pseudo partial derivative of the system at time k; Δu(k) is the input change of the system at time k; Δd(k) is the disturbance change of the system at time k;

[0021] Next, establish the control input criterion function:

[0022] J(u(k))=|y * (k+1)-y(k+1)| 2 +λ|u(k)-u(k-1)| 2 ;

[0023] Where J(u(k)) is the control input criterion function of the system at time k; λ is the weight factor; y*(k+1) is the expected output value of the system's output y(k+1) at time k+1; u(k-1) is the input of the system at time k-1;

[0024] Finally, by bringing the control input criterion function into the iterative representation of the system and taking the derivative of u(k), we can obtain the following control rate:

[0025]

[0026] Where ρ1 represents the step size factor; u(k-1) is the input of the system at time k-1.

[0027] As a further solution of the present invention, the calculation process of the total disturbance estimate is as follows:

[0028] Firstly, the linear extended state observer (LESO) is used as the total disturbance estimator for model-free adaptive control with data disturbance compensation, and the n-order nonlinear representation of the system in the time domain is constructed:

[0029]

[0030] Where y (n) is the nth-order output of the system; y (n–1) is the n-1th order output of the system; ω is the system disturbance; b is the actual gain coefficient of the system; b0 is the calibrated gain coefficient of the system; u is the input of the system; a0, a1, ..., a n-1 are the coefficients of the system equations from order 0 to n–1 respectively; f is the generalized disturbance including the internal disturbance and external disturbance of the system;

[0031] Next, the disturbance of the system is expanded into the n+1th state variable, and the state equation of the system can be described as:

[0032]

[0033] Where x is the state variable, represents the first derivative of x; represents the first-order derivative of f; A is the state matrix; B is the input matrix; E is the disturbance matrix; C is the output matrix;

[0034] The matrices are represented as follows:

[0035]

[0036] The full-order LESO state equation of the system can be expressed as:

[0037]

[0038] Where z represents the LESO output signal; represents the first derivative of z; Represents the output estimate of the system; L is the error feedback gain coefficient matrix of the system;

[0039] When L=[l1 l2…l n+1 ] T When , the characteristic polynomial of the system can be expressed as:

[0040] λ(s)=s n+1 +l1s n +…+l n s+l n+1 =(s+ω0) n+1 ;

[0041] In the formula, l1 represents the first element in L; l2 represents the second element in L; l n+1 represents the n+1th element in L; λ(s) is the characteristic polynomial of the system; s represents the wind power grid-connected system via flexible direct current, which is also the first-order representation of the system; s n+1 represents the n+1 order polynomial of the system; s n represents the nth-order polynomial of the system; ω0 is the LESO control bandwidth;

[0042] The i-th element l in L i The calculation formula is as follows:

[0043]

[0044] In the formula, i! represents the factorial of i; (n+1)! represents the factorial of n+1; (n+1-i)! represents the factorial of n+1-i;

[0045] Finally, after discretization, the iterative algorithm for the total disturbance estimate can be obtained as follows:

[0046]

[0047] Where T is the sampling period; z1(k+1), z2(k+1), z n (k+1) and z n+1 (k+1) represents the first iteration value, the second iteration value, the nth iteration value and the n+1th iteration value of z at time k+1 respectively; z1(k), z2(k), z3(k), z n (k) and z n+1 (k) represents the first iteration value, second iteration value, third iteration value, nth iteration value and n+1th iteration value of z at time k respectively;

[0048] At this point, the total disturbance estimate of the system is expressed as follows:

[0049]

[0050] Where, represents the total disturbance estimate.

[0051] As a further solution of the present invention: the calculation process of the pseudo partial derivative is as follows:

[0052] First, create the criterion function:

[0053]

[0054] Where, J(φ c (k)) represents the criterion function of the pseudo partial derivative of the system at time k; μ is the weight coefficient; It represents the estimated value of the disturbance change d(k+1) of the system at time k; is the pseudo partial derivative φ of the system at time k-1 c (k-1); Δu(k-1) is the change in the input u(k-1) of the system at time k-1; y(k-1) is the output of the system at time k-1;

[0055] Next, for the φ in the criterion function c (k) Taking the partial derivative and setting it equal to 0, we can get the estimated value of the pseudo partial derivative as follows:

[0056]

[0057] Where, is the pseudo partial derivative φ of the system at time k c (k); η is the step size factor, Δy(k) is the change in the system output y(k) at time k-1;

[0058] Finally, the pseudo partial derivative is reset, and the reset representation is as follows:

[0059]

[0060] Where, is the pseudo partial derivative φ of the system at the initial moment c (0); ε is the threshold constant.

[0061] As a further solution of the present invention: the sub-steps of step A3 are as follows:

[0062] Firstly, a model-free adaptive oscillation suppression strategy is installed in the converter control link at the sending end of the wind power flexible direct current grid-connected system;

[0063] Next, determine the parameters that need to be optimized and store them in the parameter vector r, r = [λρ1ημb0ω0] T ;

[0064] Finally, each parameter in the parameter vector is optimized. The specific optimization is as follows:

[0065]

[0066] Where, is φ c The upper bound of λ min is the lower bound of λ, and M1 is a constant greater than 0.

[0067] A model-free adaptive oscillation suppression device for a wind power grid-connected flexible DC system with a data disturbance observer, the device applying the above-mentioned model-free adaptive oscillation suppression method for a wind power grid-connected flexible DC system with a data disturbance observer, comprising:

[0068] The module for establishing the model-free adaptive oscillation suppression structure of the wind power grid-connected flexible DC system is used for the dynamic linearization of the wind power grid-connected flexible DC system and the design of the model-free adaptive oscillation suppression structure to determine the model-free adaptive oscillation suppression structure of the wind power grid-connected flexible DC system;

[0069] The model-free adaptive control algorithm design module with data disturbance compensation is used to design the controller of the model-free adaptive control with disturbance compensation, the disturbance observer of the model-free adaptive control, and the pseudo-partial derivative estimator to achieve the dynamic linearization of the wind power grid-connected flexible direct current system and the model-free adaptive oscillation suppression function;

[0070] The optimal configuration position selection of the model-free adaptive oscillation suppression strategy in the wind power flexible direct current grid-connected system and the model-free adaptive control parameter optimization module with data disturbance compensation select the optimal installation position of the model-free adaptive oscillation suppression strategy and perform the model-free adaptive control algorithm with data disturbance compensation to enable the controller to achieve the optimal control effect.

[0071] A computer device, which applies the above-mentioned model-free adaptive oscillation suppression equipment for a wind power flexible direct current grid-connected system with a data disturbance observer, comprises a processing unit and a data storage module. The data storage module is pre-installed with a program instruction set that can be called by the processing unit. When the program instruction set is called by the processing unit, each step in the method can be implemented.

[0072] A readable storage medium is applied to the above-mentioned computer device, and includes executable program code. When the executable program code is read and executed by a processing unit, the complete process of the method can be implemented.

[0073] Compared with the prior art, the present invention has the following beneficial effects:

[0074] 1. The method of this invention integrates a model-free adaptive oscillation suppression strategy with a wind power grid-connected flexible DC system to create a unique oscillation suppression structure. Leveraging a data disturbance compensation mechanism, it effectively addresses system uncertainties and improves system stability under complex operating conditions. By obtaining optimized strategy positions and parameters, it enables more precise control of wind farm operations, enhancing the reliability and safety of wind power grid connection, avoiding problems such as wind turbine downtime and equipment damage caused by oscillation, and ensuring stable power system operation.

[0075] 2. A mathematical model accurately describes the wind power grid-connected flexible direct current system, clearly defining key elements such as input, output, order, nonlinear functions, and disturbances. This provides a standardized framework for subsequent control algorithm design and system analysis. This precise mathematical representation helps researchers gain a deeper understanding of system characteristics and accurately grasp the dynamic changes in the system, providing strong support for developing targeted control strategies and solving practical engineering problems.

[0076] 3. By converting the system representation into an iterative representation and establishing a control input criterion function, this method for calculating the control rate comprehensively considers the system's desired output, input changes, and the regulatory effects of weighting factors. The resulting control rate can achieve reasonable input adjustments while maintaining system stability, bringing the system output closer to the desired state and effectively improving the system's control performance and response speed.

[0077] 4. The linear extended state observer (LESO) is used to estimate total disturbances. By constructing an nth-order nonlinear representation and state equation for the system, system disturbances are incorporated into the state variables for processing. The iterative algorithm derived from discretization can accurately estimate the total system disturbance in real time, providing a precise basis for subsequent data disturbance compensation. This improves the system's adaptability to disturbances and its anti-interference performance, and enhances the stability and reliability of system operation.

[0078] 5. By establishing a criterion function and taking the partial derivatives of the pseudo-partial derivatives to determine the estimated value, and introducing a reset mechanism, the calculation of the pseudo-partial derivatives becomes more flexible and adaptable. This calculation method can better track changes in the system's operating state and adjust the value of the pseudo-partial derivatives in a timely manner, thereby optimizing the performance of the control algorithm and improving the system's control accuracy and dynamic response capabilities under different operating conditions.

[0079] 6. The model-free adaptive oscillation suppression strategy is implemented in the control link of the sending-end converter, which is highly targeted and easy to implement. By clarifying the optimization parameters and performing reasonable optimization, the parameters can be adjusted according to the actual system operation to achieve the optimal control effect of the controller, further improving the system stability and oscillation suppression capabilities, and ensuring the efficient and stable operation of the wind power grid-connected flexible direct current system.

[0080] 7. This equipment integrates multiple functional modules, including structure establishment, algorithm design, location selection, and parameter optimization, providing an integrated solution from system structure design to control algorithm optimization. These modules work together to effectively implement dynamic linearization and model-free adaptive oscillation suppression for wind power grid-connected flexible direct current systems, providing complete and reliable technical support for practical engineering applications.

[0081] 8. The computer's pre-programmed instruction set can be called to execute the relevant method steps, enabling automated control of the model-free adaptive oscillation suppression process for the flexible direct current grid-connected wind power system. Leveraging the computer's high-speed computing and precise processing capabilities, it can quickly respond to system changes and adjust control strategies in a timely manner, improving system efficiency and accuracy, reducing manual intervention, and enhancing overall control reliability.

[0082] 9. The executable program code stored on the readable storage medium ensures the repeatability and portability of the relevant method process. The complete method process can be implemented on any computer device that supports reading the storage medium, facilitating the promotion and application of the technology and the maintenance and upgrade of the system, providing strong technical support for the widespread application of wind power grid-connected flexible direct current systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0083] Figure 1 This is a structural diagram of the wind power flexible direct current grid-connected system of the present invention.

[0084] Figure 2 Flowchart of the present invention.

[0085] Figure 3 This is a schematic diagram of the dynamic linearization of the wind power flexible direct current grid-connected system of the present invention.

[0086] Figure 4 This is a structural diagram of the model-free adaptive controller with data disturbance compensation according to the present invention.

[0087] Figure 5 This is the installation location diagram of the model-free adaptive controller with data disturbance compensation of the present invention.

[0088] Figure 6 This is a flow chart of parameter optimization of the model-free adaptive algorithm containing data disturbance compensation of the present invention.

[0089] Figure 7 This is a schematic structural diagram of the model-free adaptive oscillation suppression equipment for the wind power flexible direct current grid-connected system of the present invention.

[0090] Figure 8 Schematic diagram of the structure of the computer device of the present invention.

[0091] Figure 9 This is a graph showing the experimental results of the present invention.

[0092] In the figure: 500, computer equipment; 510, processing unit; 520, data storage module; 530, readable storage medium. DETAILED DESCRIPTION

[0093] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0094] In this embodiment, Figure 1 As shown in the figure, the wind power grid-connected system via flexible direct current transmission includes: n wind turbines connected via flexible direct current transmission system, and the system is defined as s;

[0095] like Figure 2 As shown, a model-free adaptive oscillation suppression method, equipment, device and medium for a wind power flexible direct current grid-connected system with a data disturbance observer is performed in the following steps:

[0096] Step S1: Establish a model-free adaptive oscillation suppression structure for a wind power grid-connected flexible direct current system.

[0097] Step S1.1: Dynamically linearize the wind power grid-connected flexible direct current system.

[0098] Step S1.1.1: Figure 3 As shown in Figure 2, the model-free adaptive oscillation suppression system of the wind power grid-connected flexible direct current system can be regarded as a single-input-single-output nonlinear discrete-time system, and the mathematical model of the system s can be expressed as:

[0099] y(k+1)=f(y(k),...,y(kn y ),u(k),...,u(kn u ))+d(k) (1)

[0100] In formula (1), u(k) and y(k) represent the input and output of the system at time k, respectively; y(k+1) represents the output of the system at time k+1; u(k–n u ), y(k–n y ) represent the system in k–n u The input and k–n at the moment y Output at time; n u , n y are the unknown input and output orders of the system; f(·) is the nonlinear function that characterizes the system characteristics; d(k) represents the bounded unknown disturbance to the system at time k.

[0101] Step S1.1.2: Equivalent linearization of the wind power grid-connected flexible direct current system must meet the following two assumptions:

[0102] Assumption 1: Except for a finite number of time points, the partial derivatives of the function f(·) with respect to the input u(k) are continuous.

[0103] Assumption 2: Except for a finite number of time points, the described system satisfies the generalized Lipschitz condition of the system, that is, there exists a constant b greater than 0 such that:

[0104] ∣y(k1+1)-y(k2+1)∣≤b∣u(k1)-u(k2)∣ (2)

[0105] In formula (2), y(k1+1) and y(k2+1) represent the output of the system at time k1+1 and time k2+1, respectively; u(k1) and u(k2) represent the input of the system at time k1 and time k2, respectively; k1 and k2 are constants greater than 0, and k1 is not equal to k2.

[0106] Assumption 1 is a typical constraint for general nonlinear systems; Assumption 2 indicates that bounded input changes should produce bounded system output changes, which is satisfied in most systems; the wind power grid-connected flexible direct current system is a real physical system that follows the law of conservation of energy. Limited input will only produce limited output, satisfying the conditions of Assumptions 1 and 2.

[0107] When the wind power grid-connected flexible direct current system meets assumptions 1 and 2, the wind power grid-connected flexible direct current system can be subjected to equivalent dynamic linearization; the model-free adaptive control (MFAC) equivalent dynamic linearization includes three methods, namely, the tight grid form equivalent dynamic linearization, the partial grid form equivalent dynamic linearization and the full grid form equivalent dynamic linearization; among them, the tight grid form equivalent dynamic linearization has strong real-time performance and can meet the demand for rapid response when the new energy grid-connected system oscillates, so the tight grid form dynamic linearization is adopted; the pseudo partial derivative (PPD) φ is introduced into the system c Formula (1) can be expressed as follows:

[0108] Δy(k+1)=φ c (k)Δu(k)+Δd(k) (3)

[0109] In formula (3), Δy(k+1) is the output change of the system at time k+1, Δu(k) is the input change of the system at time k, Δd(k) is the disturbance change of the system at time k, and φ c (k) is the pseudo partial derivative of the system at time k, Δy(k+1)=y(k+1)-y(k), Δu(k)=u(k)-u(k-1), Δd(k)=d(k)-d(k-1), and |φ c (k)| and Δd(k) are bounded, u(k-1) is the input of the system at time k-1, and d(k-1) represents the bounded unknown disturbance received by the system at time k-1.

[0110] At this point, the dynamic linearization of the wind power grid-connected flexible direct current system is achieved by introducing pseudo partial derivatives.

[0111] Step S1.2: Design a model-free adaptive oscillation suppression structure.

[0112] Step S1.2.1: The MFAC-based oscillation suppression strategy collects components with significant oscillation characteristics from the wind power through the flexible direct current grid-connected system, calculates the output through the MFAC system and transmits it back to the control system. The control structure consists of a pseudo partial derivative estimator, a disturbance estimator and an MFAC output device.

[0113] Step S1.2.2: The present invention includes a model-free adaptive oscillation suppression structure with data disturbance compensation as follows: Figure 4 As shown in the figure, the pseudo partial derivative estimation module adopts an improved projection algorithm to calculate the pseudo partial derivative estimation value φ of the system at the current moment by collecting real-time input and output of the system. c (k); the disturbance estimator estimates the total disturbance of the system by collecting the current output of the system, providing a basis for controlling disturbance compensation; the output of the pseudo partial derivative estimator and the output of the disturbance observer are introduced into the MFAC controller to generate a compensation signal and transmit it to the controlled system; the control signal u(k) is used as the output of the forward predictive controller and is fed back to the pseudo partial derivative estimation link as an input parameter. This cross-coupling relationship constructs a closed-loop feedback architecture, which enables the system to maintain the stability of the dynamic response process.

[0114] Step S1.2.3: MFAC oscillation suppression is different from traditional oscillation suppression methods such as active damping. MFAC has no clear physical meaning. The theoretical basis of this method is to rely on real-time input and output signals to establish a linearized virtual equivalent model of a wind power grid-connected flexible direct current system with time-varying characteristics, generate oscillation compensation through data-driven adaptive adjustment, and generate a component with a phase opposite to the oscillation component to compensate and achieve oscillation suppression; when oscillation occurs in the wind power grid-connected flexible direct current system, the voltage and current in the system will contain oscillation components; therefore, a state quantity with significant oscillation characteristics is selected, and the steady-state value of this component when no oscillation occurs is set as the tracking target of the MFAC controller, and the actual value y(k) of this state is approached to the steady-state value y* of the non-oscillating working condition by MFAC, so as to achieve oscillation suppression; in this process, the damping required for oscillation suppression is adaptively provided to the system through a data-driven iterative algorithm;

[0115] Step S2: Design a model-free adaptive control algorithm with data disturbance compensation.

[0116] Step S2.1: Design of MFAC controller with disturbance compensation.

[0117] Step S2.1.1: To facilitate algorithm design, the system described by Equation (3) is re-described into a form suitable for iterative calculation:

[0118] y(k+1)=y(k)+φ c (k)Δu(k)+Δd(k) (4)

[0119] For discrete-time systems, the control algorithm obtained by minimizing the one-step-ahead prediction error criterion function may produce excessive control input, causing damage to the controlled system itself, while the control algorithm obtained by minimizing the weighted one-step-ahead prediction error criterion function may produce steady-state tracking error. Consider the following control input criterion function:

[0120] J(u(k))=|y * (k+1)-y(k+1)| 2 +λ|u(k)-u(k-1)| 2 (5)

[0121] In formula (5), the weight factor λ>0 is used to limit the rate of change of the control input; y*(k+1) is the expected output value of the controlled system.

[0122] Step S2.1.2: To find the minimum value of J(u(k)), substitute Equation (4) into the criterion function (5) and take the derivative of u(k). The following control rate can be obtained:

[0123]

[0124] In formula (6), ρ1 represents the step size factor, which is introduced to make the algorithm more general.

[0125] Since the pseudo partial derivative φ c (k) and the system disturbance Δd(k) are both unknown terms in equation (6), so it is necessary to design an estimation algorithm to estimate their values.

[0126] Step S2.2: Model-free adaptive control disturbance observer design.

[0127] Step S2.2.1: The wind power grid-connected flexible direct current system contains a large number of fully controlled power electronic devices. The switching action will introduce nonlinear disturbance signals into the system. At the same time, there are external noise disturbances. In order to improve the robustness of the control system, the system disturbance needs to be compensated. Common data disturbance compensation methods include filtering methods, transform domain methods, etc. This patent introduces an observer to perform data disturbance compensation.

[0128] To estimate the total disturbance of the system, a Linear Expansion State Observer (LESO) is used. LESO can accurately estimate internal and external disturbances in real time and can be used as a total disturbance estimator for Model-Free Adaptive Control with Data Perturbation Compensation (DPC-MFAC). An n-order nonlinear system can be described in the time domain as:

[0129]

[0130] In formula (7): u is the system input, y (n) is the nth-order output of the system, y (n–1) is the n-1th order output of the system, ω is the system disturbance, b is the system gain coefficient, a0, a1, …, a n-1 are the coefficients of order 0 to n–1 of the system equations, respectively.

[0131] The actual system gain b is difficult to determine. Using b0 can approximately replace the system gain, and the n-order equation of the system can be expressed as:

[0132]

[0133] In formula (8), f is the generalized disturbance that includes the internal disturbance and external disturbance of the system.

[0134] Step S2.2.2: Expand the system disturbance into the n+1th state variable. The state equation of the system can be described as:

[0135]

[0136] Among them, the state matrix A and input matrix B are:

[0137]

[0138] The perturbation matrix E and output matrix C are:

[0139]

[0140] At this time, the system described by Equation (7) introduces n+1 state variables, and the full-order LESO state equation of the system can be expressed as:

[0141]

[0142] Where, LESO output signal z=[z1 z2 … z n+1 ] T, L is the error feedback gain coefficient matrix of the system, when L=[l1 l2 … l n+1 ] T When , the characteristic polynomial of the system can be expressed as:

[0143]

[0144] Among them, l i It can be expressed as:

[0145]

[0146] By discretizing the above, we can get the perturbation iterative algorithm as follows:

[0147]

[0148] In formula (15), T is the sampling period. When T is small enough and the control bandwidth is large enough, the observer disturbance estimation error converges exponentially.

[0149] At this point, the total disturbance estimate of the system can be expressed as:

[0150]

[0151] At this point, the introduction of the LESO observer realizes the estimation of the total disturbance of the system and can provide compensation for the disturbance suppression of the controller.

[0152] Step S2.3: Pseudo-partial derivative estimator design.

[0153] Step S2.3.1: For the oscillation suppression of wind power system connected to the flexible DC grid, based on the improved forward prediction algorithm, the pseudo-partial derivative of the controller can be established by establishing the criterion function J(φ c (k)) Make an estimate:

[0154]

[0155] Where μ is the weight coefficient, which is used to punish excessive changes in PPD valuation. is the estimated value of Δd, is φ c estimated value of;

[0156] Step S2.3.2: Equation (17) for φ c (k) Taking the partial derivative and setting the derivative equal to 0, we can get the improved PPD estimation algorithm:

[0157]

[0158] In formula (18), η∈(0, 1) is the step size factor, which is introduced to make the PPD estimation algorithm more general.

[0159] In order to enhance the algorithm's tracking of time-varying running states, the following reset algorithm is introduced:

[0160]

[0161] In formula (19), is the initial value of PPD, and ε is a very small constant.

[0162] Step S3: Obtain the optimal configuration position of the model-free adaptive oscillation suppression strategy in the wind power flexible direct current grid-connected system and perform parameter optimization.

[0163] Step S3.1: Obtain the installation location of the improved MFAC oscillation suppression.

[0164] Step S3.1.1: In the wind power flexible direct current grid-connected projects that have been put into operation, due to the large number of wind turbines and their dispersed distribution, it is difficult to reconstruct the control of each wind turbine once it is put into operation. For the oscillation suppression of wind power flexible direct current grid-connected projects, additional equipment is usually selected or additional control is performed on the flexible direct current sending end converter station; therefore, the DPC-MFAC oscillation suppression is installed in the control link of the converter at the sending end of the flexible direct current transmission system; DPC-MFAC has strong adaptability, and the location of additional control can be selected in a large range; the flexible direct current transmission system adopts proportional integral control in the dq rotating coordinate system. When oscillation occurs, the actual value of the d-axis PI control deviates from the reference value, generating a component with the same frequency as the oscillation; therefore, the input parameter of the modified DPC-MFAC oscillation suppression can be selected as the deviation between the d-axis current reference value and the actual value of the MMC converter, and the improved MFAC output is introduced into the d-axis feedforward link, such as Figure 5 As shown;

[0165] When the system is running stably, the MFAC output is 0. When the oscillation monitoring system detects system oscillation, the MFAC oscillation suppression strategy is activated.

[0166] Step S3.2: Improve MFAC oscillation suppression parameter tuning and optimization.

[0167] Step S3.2.1: Although the improved MFAC oscillation suppression algorithm has strong adaptive capabilities, the control performance and response speed of the system can be improved by properly selecting parameters. Therefore, it is necessary to optimize the control parameters. Although MFAC has derived a variety of improved algorithms, its parameter tuning process still lacks systematic guiding principles. The initial value of the DPC-MFAC pseudo partial derivative φ c (0) It has strong adaptive ability and can be selected in a wide range; the parameters that need to be optimized in the DPC-MFAC algorithm are r = [λ ρ1 η μ b0 ω0] T Based on the assumptions of stability proof, certain constraints are imposed on the selection of system parameters:

[0168]

[0169] In formula (20), is φ c The upper bound of λ min is the lower bound of λ, and M1 is a constant greater than 0.

[0170] The parameter design principles are as follows:

[0171] (1) The smaller λ is, the faster the system responds. However, too fast a response may cause overshoot.

[0172] (2) The step factor ρ1 affects the output change rate of the system. The larger its value, the greater the output change of the controller in each step.

[0173] (3) The step size factor η of PPD affects the rate of change of PPD. The larger η is, the smaller the amplitude of the controlled system is.

[0174] (4) The weight coefficient μ of the controller PPD performance index function determines the constraint strength of PPD. The smaller the amplitude of the controlled system, the greater the μ. If μ is too small, the system may become unstable. If it is too large, the system oscillation amplitude will increase and the convergence speed will slow down.

[0175] (5) The larger the disturbance observer control bandwidth ω0, the faster the system responds, but too high ω0 may cause overshoot and oscillation.

[0176] (6) The estimated value of the system gain b0 should be as close as possible to the actual control gain b of the system; if the model is known, the nominal value can be directly taken; if it is unknown, the system can be identified and estimated through step response, etc.

[0177] Step S3.2.2: The above parameter setting method is relatively rough. In order to further improve the system stability and response speed, time domain simulation can be used to optimize the parameters by using performance index functions and combining external optimization algorithms. In view of the oscillation characteristics of the wind power grid-connected flexible direct current system, the product of the input variable containing the oscillation signal and the time is selected as the performance index, and the parameters are optimized through PSCAD / EMTDC multiple runtime time domain simulation combined with external MATLAB optimization algorithm.

[0178] like Figure 6As shown in the figure, an oscillation suppression model of a wind power grid-connected flexible direct current system including DPC-MFAC is built in PSCAD / EMTDC, and values ​​are assigned according to the constraints, and the PSCAD / EMTDC simulation process is started. After the simulation process is completed, the optimization program automatically calls the generated ".out" file and calculates the system performance indicators through an external optimization program. If the parameters meet the preset accuracy requirements, the operation is stopped and the optimal parameters are given; if not, the system parameters are reassigned and a new round of PSCAD / EMTDC simulation program is run until the optimal solution that meets the system performance indicators is obtained.

[0179] Step S4: A model-free adaptive oscillation suppression device for a wind power grid-connected flexible direct current system with a data disturbance observer, such as Figure 7 As shown, it is characterized in that the method described in any one of claims S1 to S3 is used, including: a module for establishing a model-free adaptive oscillation suppression structure of a wind power flexible direct current grid-connected system, which is used for dynamic linearization of the wind power flexible direct current grid-connected system and design of a model-free adaptive oscillation suppression structure to determine the DPC-MFAC structure of the wind power flexible direct current grid-connected system.

[0180] The model-free adaptive control algorithm design module with data disturbance compensation designs the MFAC controller with disturbance compensation, the model-free adaptive control disturbance observer and the pseudo-partial derivative estimator respectively, so as to realize the dynamic linearization of the wind power flexible direct current grid-connected system and the model-free adaptive oscillation suppression function.

[0181] The optimal configuration location selection and DPC-MFAC parameter optimization module of the model-free adaptive oscillation suppression strategy in the wind power flexible direct current grid-connected system selects the optimal installation location of the model-free adaptive oscillation suppression strategy and performs the DPC-MFAC algorithm to enable the controller to achieve the optimal control effect.

[0182] Step S5: Please refer to the diagram of the structure of the computer device in the embodiment of the present application. Figure 8 ;The computer device 500 provided in this embodiment mainly includes a processing unit 510 and a data storage module 520, wherein the data storage module 520 is pre-installed with a program instruction set that can be called by the processing unit 510. When the program instruction set is called by the processing unit 510, the various implementation steps of the technical solution described above can be implemented.

[0183] Step S6: The technical solution of the present application also provides a non-transitory computer-readable storage medium 530, which records executable program code. When the program code is read and executed by the processing unit 510, the complete process of the method described above can be implemented.

[0184] It should be noted that the storage medium 530 can be implemented by a combination of various volatile or non-volatile storage devices, specifically including but not limited to the following types: static random access memory SRAM, electrically rewritable programmable read-only memory EEPROM, erasable programmable read-only memory EPROM, one-time programmable read-only memory PROM, fixed read-only memory ROM, magnetic storage devices, flash memory storage devices, and various optical or magnetic disk storage media.

[0185] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with the technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solution and inventive concept of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. A model-free adaptive oscillation suppression method for a wind power grid-connected flexible direct current system with a data disturbance observer, characterized in that: The following steps are involved: A1. Based on the model-free adaptive oscillation suppression strategy and the wind power grid-connected flexible DC system, a model-free adaptive oscillation suppression structure for the wind power grid-connected flexible DC system is established; A2. Compensate for data disturbances in the wind power grid-connected flexible direct current system by calculating the control rate, total disturbance estimate, and pseudo-partial derivative including disturbance compensation. A3. Obtain the position of the model-free adaptive oscillation suppression strategy in the wind power flexible direct current grid-connected system after data disturbance compensation, and perform corresponding parameter optimization; A4. The operation of wind farms is controlled by the flexible direct current grid-connected system with optimized wind power parameters.

2. The method for model-free adaptive oscillation suppression of a wind power grid-connected flexible direct current system with a data disturbance observer according to claim 1, characterized in that: The wind power connected to the grid through flexible direct current system is represented as follows: y(k+1)=f(y(k),...,y(k-n y ),u(k),...,u(k-n u ))+d(k); Where u(k) and y(k) are the input and output of the system at time k, respectively; y(k+1) is the output of the system at time k+1; u(k–n u ) and y(k–n y ) are the system in k–n u The input and k–n at the moment y Output at time; n u and n y are the input order and output order of the system respectively; f(·) is the nonlinear function that characterizes the characteristics of the system; d(k) is the bounded unknown disturbance received by the system at time k.

3. The method for model-free adaptive oscillation suppression of a wind power grid-connected flexible direct current system with a data disturbance observer according to claim 2, characterized in that: The calculation process of control rate is as follows: First, the system representation is converted into an iterative representation that can be iteratively calculated. The specific representation is as follows: y(k+1)=y(k)+φ c (k)Δu(k)+Δd(k); Where, φ c (k) is the pseudo partial derivative of the system at time k; Δu(k) is the input change of the system at time k; Δd(k) is the disturbance change of the system at time k; Next, establish the control input criterion function: J(u(k))=|y * (k+1)-y(k+1)| 2 +λ|u(k)-u(k-1)| 2 ; Where J(u(k)) is the control input criterion function of the system at time k; λ is the weight factor; y*(k+1) is the expected output value of the system's output y(k+1) at time k+1; u(k-1) is the input of the system at time k-1; Finally, by bringing the control input criterion function into the iterative representation of the system and taking the derivative of u(k), we can obtain the following control rate: Where ρ1 represents the step size factor; u(k-1) is the input of the system at time k-1.

4. The method for model-free adaptive oscillation suppression of a wind power grid-connected flexible direct current system with a data disturbance observer according to claim 3 is characterized in that: The total disturbance estimate is calculated as follows: Firstly, the linear extended state observer (LESO) is used as the total disturbance estimator for model-free adaptive control with data disturbance compensation, and the n-order nonlinear representation of the system in the time domain is constructed: Where y (n) is the nth-order output of the system; y (n–1) is the n-1th order output of the system; ω is the system disturbance; b is the actual gain coefficient of the system; b0 is the calibrated gain coefficient of the system; u is the input of the system; a0, a1, ..., a n-1 are the coefficients of the system equations from order 0 to n–1 respectively; f is the generalized disturbance including the internal disturbance and external disturbance of the system; Next, the disturbance of the system is expanded into the n+1th state variable, and the state equation of the system can be described as: Where x is the state variable, represents the first derivative of x; represents the first-order derivative of f; A is the state matrix; B is the input matrix; E is the disturbance matrix; C is the output matrix; The matrices are represented as follows: The full-order LESO state equation of the system can be expressed as: Where z represents the LESO output signal; represents the first derivative of z; Represents the output estimate of the system; L is the error feedback gain coefficient matrix of the system; When L=[l1 l2 … l n+1 ] T When , the characteristic polynomial of the system can be expressed as: λ(s)=s n+1 +l1s n +…+l n s+l n+1 =(s+ω0) n+1 ; In the formula, l1 represents the first element in L; l2 represents the second element in L; l n+1 Represents the n+1th element in L; λ(s) is the characteristic polynomial of the system; s represents the wind power grid-connected system through flexible direct current, which is also the first-order representation of the system; s n+1 represents the n+1 order polynomial of the system; s n represents the nth-order polynomial of the system; ω0 is the LESO control bandwidth; The i-th element l in L i The calculation formula is as follows: In the formula, i! represents the factorial of i; (n+1)! represents the factorial of n+1; (n+1-i)! represents the factorial of n+1-i; Finally, after discretization, the iterative algorithm for the total disturbance estimate can be obtained as follows: Where T is the sampling period; z1(k+1), z2(k+1), z n (k+1) and z n+1 (k+1) represents the first iteration value, the second iteration value, the nth iteration value and the n+1th iteration value of z at time k+1 respectively; z1(k), z2(k), z3(k), z n (k) and z n+1 (k) represents the first iteration value, second iteration value, third iteration value, nth iteration value and n+1th iteration value of z at time k respectively; At this point, the total disturbance estimate of the system is expressed as follows: Where, represents the total disturbance estimate.

5. The method for model-free adaptive oscillation suppression of a wind power flexible direct current grid-connected system with a data disturbance observer according to claim 4 is characterized in that: The calculation process of pseudo partial derivatives is as follows: First, create the criterion function: Where, J(φ c (k)) represents the criterion function of the pseudo partial derivative of the system at time k; μ is the weight coefficient; It represents the estimated value of the disturbance change d(k+1) of the system at time k; is the pseudo partial derivative φ of the system at time k-1 c The estimated value of (k-1); Δu(k-1) is the change in the system input u(k-1) at time k-1; y(k-1) is the system output at time k-1; Next, for the φ in the criterion function c (k) Taking the partial derivative and setting it equal to 0, we can get the estimated value of the pseudo partial derivative as follows: Where, is the pseudo partial derivative φ of the system at time k c (k); η is the step size factor, Δy(k) is the change in the system output y(k) at time k-1; Finally, the pseudo partial derivative is reset, and the reset representation is as follows: Where, is the pseudo partial derivative φ of the system at the initial moment c (0); ε is the threshold constant.

6. The method for model-free adaptive oscillation suppression of a wind power grid-connected flexible direct current system with a data disturbance observer according to claim 5, characterized in that: The sub-steps of step A3 are as follows: Firstly, a model-free adaptive oscillation suppression strategy is installed in the converter control link at the sending end of the wind power flexible direct current grid-connected system; Next, determine the parameters that need to be optimized and store them in the parameter vector r, r = [λρ1ημb0ω0] T ; Finally, each parameter in the parameter vector is optimized. The specific optimization is as follows: Where, is φ c The upper bound of λ min is the lower bound of λ, and M1 is a constant greater than 0.

7. A model-free adaptive oscillation suppression device for a wind power grid-connected flexible DC system with a data disturbance observer, the device applying a model-free adaptive oscillation suppression method for a wind power grid-connected flexible DC system with a data disturbance observer according to any one of claims 1 to 6, characterized in that: include: The module for establishing the model-free adaptive oscillation suppression structure of the wind power grid-connected flexible DC system is used for the dynamic linearization of the wind power grid-connected flexible DC system and the design of the model-free adaptive oscillation suppression structure to determine the model-free adaptive oscillation suppression structure of the wind power grid-connected flexible DC system; The model-free adaptive control algorithm design module with data disturbance compensation is used to design the controller of the model-free adaptive control with disturbance compensation, the disturbance observer of the model-free adaptive control, and the pseudo-partial derivative estimator to achieve the dynamic linearization of the wind power grid-connected flexible direct current system and the model-free adaptive oscillation suppression function; The optimal configuration position selection of the model-free adaptive oscillation suppression strategy in the wind power flexible direct current grid-connected system and the model-free adaptive control parameter optimization module with data disturbance compensation select the optimal installation position of the model-free adaptive oscillation suppression strategy and perform the model-free adaptive control algorithm with data disturbance compensation to enable the controller to achieve the optimal control effect.

8. A computer device, which uses the model-free adaptive oscillation suppression equipment for wind power grid-connected flexible direct current systems with a data disturbance observer as claimed in claim 7, characterized in that: The method comprises a processing unit (510) and a data storage module (520). The data storage module (520) is pre-installed with a program instruction set that can be called by the processing unit (510). When the program instruction set is called by the processing unit (510), each step in the method can be implemented.

9. A readable storage medium, applied to the computer device according to claim 8, characterized in that: The method comprises an executable program code, which can realize the complete process of the method when the executable program code is read and executed by the processing unit (510).

Images