Energy storage converter sliding mode active-disturbance-rejection control method based on improved ESO
By designing an improved fourth-order nonlinear expansion state observer and complementary superspiral sliding mode feedback control in the energy storage converter, combined with the voltage outer ring-current inner ring control, the DC bus voltage fluctuation and System stability issues have achieved more efficient power quality and dynamic performance improvement.
Patent Information
- Application Number
- CN202410528242.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-04-29
- Publication Date
- 2024-11-26
AI Technical Summary
When the energy storage converter faces load switching, random power direction switching and external disturbance, the DC bus voltage fluctuates severely, affecting the system stability. It is difficult for traditional linear control methods to quickly match the load, suppress voltage fluctuations, and non-linear control. Strategies such as ADRC and SMC have insufficient response speed and accuracy, and the vibration of SMC affects system stability.
Designed based on the improved fourth-order nonlinear expansion state observer (IFNLESO) and complementary superspiral sliding mode feedback control law, combining the dual-loop control of the voltage outer ring-current inner ring to enhance disturbance immunity, improve response speed and accuracy, and pass Super spiral sliding mode control suppresses jitter, achieving accurate compensation and rapid response to lumped disturbances.
It significantly improves the output power quality and dynamic performance of the energy storage converter, suppresses DC bus voltage fluctuations, enhances system stability and robustness, reduces harmonics on the AC side, and improves power response speed and voltage regulation efficiency.
Smart Images

Figure CN119029965A_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the technical field of power electronic control, and in particular relates to a sliding mode anti-disturbance control method for an energy storage converter based on an improved ESO. Background Art
[0002] Due to the uncertainty and intermittency of renewable energy generation, it will seriously affect the safe and stable operation of the system. Among them, the energy storage system has the advantages of storing energy on demand, which can effectively smooth out power fluctuations, improve system stability, and also has the functions of peak shaving and valley filling, and improving power quality, so it has been widely studied.
[0003] The power conversion system (PCS) is the medium for energy flow between the energy storage system and the AC power grid, and is the most important part of the microgrid grid-connected system. Therefore, in order to improve the power consumption capacity, reduce grid-connected harmonics, suppress fluctuations, and ensure the stable operation of the energy storage microgrid, the PCS control strategy has become a hot research issue. Since PCS will face complex working conditions such as load switching, random switching of power direction, and external disturbances in actual applications, this will inevitably cause the DC bus voltage to fluctuate and affect the stability of the system. Therefore, how to improve the control performance of PCS to effectively smooth out voltage fluctuations, improve the bidirectional flow and matching load capacity of energy storage and release, and increase the stability, safety and economy of the system is of great significance. The control strategies of the PCS bidirectional DC / AC link include voltage and current dual closed-loop control, constant power control, constant voltage and frequency control, and droop control. At present, most traditional PCSs use dual closed-loop control to control the DC bus voltage to a constant value to reduce the impact of the inrush current on the energy storage battery.
[0004] Since PCS is a time-varying nonlinear system with strong coupling, it is difficult for traditional linear control methods to achieve the requirements of fast load matching and suppressing the DC bus voltage fluctuations caused by large load power changes. For such nonlinear systems, nonlinear control can achieve good control requirements. However, there are few research reports on the application of nonlinear control strategies such as Active Disturbance Rejection Control (ADRC) and Sliding Mode Control (SMC) to PCS.
[0005] ADRC has outstanding advantages in complex and high-order systems due to its natural decoupling and independence from the precise model of the object. Some literature applies ADRC to the control of photovoltaic energy storage bidirectional DC-DC converters, using the traditional third-order linear extended state observer (ESO) to accurately observe the actual inductor current and internal and external disturbances, effectively suppressing the DC bus voltage fluctuations and shocks, and improving the anti-interference ability of the photovoltaic energy storage system; some literature introduces the ESO in ADRC into the second-order disturbance compensation to improve the tracking accuracy and anti-interference performance of the system. However, the above ESO still has problems such as insufficient response speed and insufficient accuracy.
[0006] SMC is slightly affected by external disturbances and inaccurate modeling, and can overcome system uncertainties and has strong robustness. However, the jitter in SMC will affect the stability of the system. Summary of the invention
[0007] The purpose of the present invention is to provide a sliding mode anti-disturbance control method for an energy storage converter based on an improved ESO, so as to improve the output power quality and dynamic performance of the energy storage converter.
[0008] The technical solution adopted by the present invention is to adopt a sliding mode anti-disturbance control method for energy storage converters based on improved ESO, adopt a dual-loop control combining a voltage outer loop and a current inner loop for the energy storage converter, and design CSTSM-ADRC as a controller of a voltage outer loop and a current inner loop to control a bidirectional AC / DC converter to stabilize the DC bus voltage and realize energy conversion between the DC side and the AC side; the bidirectional DC / DC converter adopts a PI current single-loop control strategy, and controls the output power of the energy storage battery through the bidirectional DC / DC converter to match the power required by the load;
[0009] The design of CSTSM-ADRC is as follows: first, the mathematical model of the converter is fitted as a second-order active disturbance rejection model, and an improved fourth-order nonlinear extended state observer is designed to achieve more accurate estimation and compensation of the lumped disturbance and enhance the anti-disturbance capability. The response speed is improved by using nonlinear functions, and the observer parameters are optimized. Secondly, a complementary super-helical sliding mode feedback control law is designed, and the complementary sliding surface and the super-helical sliding mode algorithm are combined as the switching law to effectively suppress the chattering.
[0010] The present invention is also characterized in that:
[0011] The CSTSM-ADRC design is implemented in the following steps:
[0012] Step 1: Establish PCS model;
[0013] Step 2: Simplify the adjustable parameters of the observer based on the ADRC principle;
[0014] Step 3: IFNLESO design;
[0015] Step 4: Design of CSTSMC feedback control law.
[0016] Step 1 is specifically as follows:
[0017] Under three-phase voltage balance, the PCS topology establishes a mathematical model in a three-phase stationary coordinate system described by the switching function:
[0018]
[0019] In formula (1), S j is the switching function, defined as:
[0020]
[0021] The PCS mathematical model in the d and q two-phase rotating coordinate system is obtained through Park transformation:
[0022]
[0023] In formula (2), v d 、v q Respectively represent the components of the AC grid voltage on the d and q axes; i d 、i q are the components of the AC grid-side current on the d and q axes respectively; S d , S q are the electrical switching functions on the d and q axes on the grid side respectively; ω is the grid voltage angular frequency.
[0024] Step 2 is specifically as follows:
[0025] In order to achieve the unity power factor of the unit and connect to the grid, the reactive current is set to zero. From equation (3), we know that the mathematical model of the bidirectional AC / DC converter is a second-order system. We take the derivative of equation (3) and fit the voltage outer loop and the d-axis and q-axis current inner loops as second-order systems containing disturbances. The state variable of the voltage loop system is defined as x v1 =v d , Current loop system state variable x d1 =i d , x q1 =i q , The obtained voltage loop second-order system is:
[0026]
[0027] The second-order current loop system is:
[0028]
[0029] In formula (4) and formula (5), b d and b q are the control gains of the d-axis and q-axis respectively; the internal influence F vd , external disturbance F d and the lumped disturbance F of various unknown inaccurate models q for:
[0030]
[0031] For the second-order system design shown in Equation (4) and Equation (5), the traditional linear ADRC is designed as a second-order system including disturbances. The second-order system is as follows:
[0032]
[0033] In formula (7), u is the input; y is the system output; is the lumped disturbance, and b in equation (7) is replaced by the control gain estimate b0 which is easy to obtain;
[0034] The three second-order systems have the same form;
[0035] Select state variable x1=y, Then the system shown in the above equation can be written as:
[0036]
[0037] Based on formula (8), the third-order linear ESO is constructed as follows:
[0038]
[0039] In formula (9), z1, z2 and z3 are the output y and the output change rate. and the tracking signal of the lumped disturbance f; ξ1, ξ2 and ξ3 are the error correction gains of the ESO;
[0040] The traditional linear SEF is constructed as:
[0041] u0=k p (y-z1)-k d z2 (10)
[0042] In formula (10), k p and k d is the controller parameter;
[0043] The lumped disturbance compensation link is designed as follows:
[0044]
[0045] According to the bandwidth method of linear observer, all poles of ESO are configured to ω c At this point, the following configuration is made for equation (8):
[0046]
[0047] The above configuration simplifies the three adjustable parameters of the observer to one.
[0048] Step 3 is as follows:
[0049] Improvements are made to the traditional linear third-order ESO by adding a state variable z4, which is the rate of change of the lumped disturbance, to observe the changing trend of the lumped disturbance, make effective compensation in advance, improve the observation accuracy, and enhance the anti-disturbance capability. At the same time, in order to increase the observation speed, a nonlinear function is added to the linear observer to improve the convergence speed and accuracy of the ESO. Therefore, a nonlinear fal function is added on the basis of the fourth-order linear ESO to improve the performance of the ESO.
[0050] The rate of change of the lumped disturbance of system (8) is Expanded to x4, the system is reconstructed as follows:
[0051]
[0052] The fourth-order linear ESO equations constructed for equation (13) are:
[0053]
[0054] In formula (14), z1, z2, z3 and z4 are output y, output change rate The aggregate disturbance f and the rate of change of the aggregate disturbance The tracking signal; ξ1, ξ2, ξ3 and ξ4 are the error gains of ESO;
[0055] Assume that in equation (14) is a constant ω F , the fourth-order linear ESO observation error is obtained as:
[0056]
[0057] When the error system enters the steady state, the steady-state error |e1|=ω is obtained. F / ξ4,|e2|=ξ1ω F / ξ4,|e3|=ξ2ω F / ξ4,|e4|=ξ3ω F / ξ4;
[0058] In order to further accelerate the convergence speed of the observer and improve the accuracy, and enhance the observation ability of the observer, an IFNLESO is proposed for the system shown in equation (13):
[0059]
[0060] In formula (16), α1=3 / 4, α2=2 / 4, α2=1 / 4, and the nonlinear function fal is as follows:
[0061]
[0062] In formula (17), δ is the error bound;
[0063] The state error equation obtained from equations (13) and (16) is:
[0064]
[0065] And let fal=|e| a signe(), when the error system enters the steady state, the steady-state errors of the observer are |e1|=(ω F / ξ4) 4 , e2=ξ3(ω F / ξ4) 4 , e3=ξ2(ω F / ξ4) 3 , e4=ξ3(ω F / ξ4) 2 .
[0066] Step 4 is as follows:
[0067] The error state equation is constructed by equations (10) and (12):
[0068]
[0069] Define the generalized sliding surface s g (t) is:
[0070]
[0071] In formula (32), β is the sliding surface constant, and β satisfies > 0;
[0072] Design a complementary sliding surface s that is orthogonal to the generalized sliding surface c (t) is:
[0073]
[0074] According to s g and c Get the total sliding surface s m :
[0075] s m =s g +sc =2ε(34)
[0076] Determine g and c The relationship between derivatives:
[0077]
[0078] Choose a Lyapunov function:
[0079]
[0080] Taking the derivative of Lyapunov, we get:
[0081]
[0082] Under the condition that the Lyapunov function satisfies the stability condition, the complementary sliding mode control law is obtained:
[0083] u t =u eq +u rl (38)
[0084] In formula (38), u eq Expressed as the sliding mode equivalent control law:
[0085]
[0086] Taking the super-helical sliding mode control law as the CSTSMC switching control law, then u in equation (33) rl Designed for:
[0087]
[0088] In formula (40), sign(·) is the sign function, and ρ and κ are positive real numbers.
[0089] The beneficial effects of the present invention are:
[0090] The present invention is based on an improved ESO sliding mode auto-disturbance rejection control method for an energy storage converter, and introduces an increased-order filtered super-helical ESO into the auto-disturbance rejection control, which can achieve more accurate and faster compensation for the lumped disturbance when a large estimation error occurs, and improve the ability to suppress the cross-influence of the q and d axes. The observer has a simple structure, does not have additional parameters, and is easy to implement in engineering. The super-helical feedback control law replaces the sliding mode switching control part in the traditional complementary sliding mode control, which can ensure that the sliding mode stage and the arrival stage can converge more quickly, further eliminate the jitter caused by high-frequency switching, and enhance the stability and robustness of the system. BRIEF DESCRIPTION OF THE DRAWINGS
[0091] Figure 1 It is the block diagram of PCS control strategy structure;
[0092] Figure 2 It is the PCS circuit topology diagram;
[0093] Figure 3 It is the ADRC schematic diagram;
[0094] Figure 4 It is the block diagram of IFNLESO structure;
[0095] Figure 5 It is the CSTSMC strategy diagram;
[0096] Figure 6 It is a diagram of the hardware-in-the-loop test system;
[0097] FIG7 is an experimental waveform diagram of different controls at a mutation of -30KW to -70KW, wherein FIG7(a) is a waveform diagram of the transient response of the DC bus voltage and the transient response of the PCS AC side voltage and current under the PI control strategy, FIG7(b) is a waveform diagram of the transient response of the DC bus voltage and the transient response of the PCS AC side voltage and current under the traditional ADRC strategy, and FIG7(c) is a waveform diagram of the transient response of the DC bus voltage and the transient response of the PCS AC side voltage and current under the control method of the present invention;
[0098] FIG8 is an experimental waveform diagram of different controls at a mutation from -30KW to 30KW, wherein FIG8(a) is a waveform diagram of the transient response of the DC bus voltage and the transient response of the PCS AC side voltage and current under the PI control strategy, FIG8(b) is a waveform diagram of the transient response of the DC bus voltage and the transient response of the PCS AC side voltage and current under the traditional ADRC strategy, and FIG8(c) is a waveform diagram of the transient response of the DC bus voltage and the transient response of the PCS AC side voltage and current under the control method of the present invention;
[0099] FIG9 is an experimental waveform diagram of different controls at a mutation of -30KW to -70KW, wherein FIG9(a) is a waveform diagram of the transient response of the DC bus voltage and the transient response of the PCS AC side voltage and current under the PI control strategy, FIG9(b) is a waveform diagram of the transient response of the DC bus voltage and the transient response of the PCS AC side voltage and current under the traditional ADRC strategy, and FIG9(c) is a waveform diagram of the transient response of the DC bus voltage and the transient response of the PCS AC side voltage and current under the control method of the present invention;
[0100] FIG10 is an experimental waveform diagram of different controls at a sudden change from 30KW to -30KW, wherein FIG10(a) is a waveform diagram of the transient response of the DC bus voltage and the transient response of the PCS AC side voltage and current under the PI control strategy, FIG10(b) is a waveform diagram of the transient response of the DC bus voltage and the transient response of the PCS AC side voltage and current under the traditional ADRC strategy, and FIG10(c) is a waveform diagram of the transient response of the DC bus voltage and the transient response of the PCS AC side voltage and current under the control method of the present invention;
[0101] FIG11 is a diagram of the transient change of the DC bus voltage and the power response time when the load power fluctuates under different control strategies, wherein FIG11(a) is a diagram of the DC bus voltage overshoot, FIG11(b) is a diagram of the DC bus voltage regulation time, and FIG11(c) is a diagram of the power response time. DETAILED DESCRIPTION
[0102] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments.
[0103] Example 1
[0104] This embodiment provides a sliding mode anti-disturbance control method for an energy storage converter based on an improved ESO. Figure 1 As shown, a dual-loop control combining a voltage outer loop and a current inner loop is adopted for the energy storage converter, and a complementary super-twisting sliding mode anti-disturbance control CSTSM-ADRC (Complementary Super-Twisting Sliding Mode ADRC) is designed as a controller of the voltage outer loop and the current inner loop to control the bidirectional AC / DC converter to stabilize the DC bus voltage and realize the energy conversion between the DC side and the AC side; the bidirectional DC / DC converter adopts a PI current single-loop control strategy, and the output power of the energy storage battery is controlled by the bidirectional DC / DC converter to match the power required by the load;
[0105] The design of CSTSM-ADRC is as follows: first, the mathematical model of the converter is fitted as a second-order active disturbance rejection model, and an improved fourth-order nonlinear extended state observer is designed to achieve more accurate estimation and compensation of the lumped disturbance and enhance the anti-disturbance capability. The response speed is improved by using nonlinear functions, and the observer parameters are optimized. Secondly, a complementary super-helical sliding mode feedback control law is designed, and the complementary sliding surface and the super-helical sliding mode algorithm are combined as the switching law to effectively suppress the chattering.
[0106] Example 2
[0107] This embodiment provides a sliding mode active disturbance rejection control method for an energy storage converter based on an improved ESO. On the basis of Embodiment 1, the CSTSM-ADRC design is specifically implemented according to the following steps:
[0108] Step 1: Establish PCS model;
[0109] like Figure 2 The circuit topology of PCS is shown in Figure 1, which can realize bidirectional power flow and reliable grid connection of energy storage system. dc is the DC voltage measurement, C is the DC capacitance measurement, R is the equivalent resistance of the grid side line, L is the filter inductance, L dc DC measurement of bidirectional DC / DC converter inductance, i a 、i b 、i c is the PCS output current, u a 、u b 、u c Measure the voltage of each phase for PCS AC, e a 、e b 、e c is the grid voltage, T1-T6 are the switch tubes of the bidirectional converter;
[0110] Under three-phase voltage balance, the PCS topology establishes a mathematical model in a three-phase stationary coordinate system described by the switching function:
[0111]
[0112] In formula (1), S j is the switching function, defined as:
[0113]
[0114] The PCS mathematical model in the d and q two-phase rotating coordinate system is obtained through Park transformation:
[0115]
[0116] In formula (2), v d 、v q Respectively represent the components of the AC grid voltage on the d and q axes; i d 、i q are the components of the AC grid-side current on the d and q axes respectively; S d , S q are the electrical switching functions on the d and q axes on the grid side respectively; ω is the grid voltage angular frequency.
[0117] Step 2: Simplify the adjustable parameters of the observer based on the ADRC principle;
[0118] The traditional second-order linear ADRC principle is as follows Figure 3As shown in the figure, it consists of three parts: Tracking Differentiator (TD), ESO, and Nonlinear State Error Feedback (NLSEF). The role of TD is to soften the start-up link signal and eliminate the conflict between overshoot and rapidity. In order to avoid high-frequency oscillation of the bus voltage, TD is not used. Figure 3 In it: v is the output reference; b0 is the control gain; u is the control quantity, z1, z2, z3 are the object output, change rate and lumped disturbance observed by ESO; y is the object output.
[0119] In order to achieve the unity power factor of the unit and connect to the grid, the reactive current is set to zero. From equation (3), we know that the mathematical model of the bidirectional AC / DC converter is a second-order system. We take the derivative of equation (3) and fit the voltage outer loop and the d-axis and q-axis current inner loops as second-order systems containing disturbances. The state variable of the voltage loop system is defined as x v1 =v d , Current loop system state variable x d1 =i d , x q1 =i q , The obtained voltage loop second-order system is:
[0120]
[0121] The second-order current loop system is:
[0122]
[0123] In equations (4) and (5), b is the control gain; the internal influence F vd , external disturbance F d and the lumped disturbance F of various unknown inaccurate models q for:
[0124]
[0125] For the second-order system design shown in Equation (4) and Equation (5), the traditional linear ADRC is designed as a second-order system including disturbances. The second-order system is as follows:
[0126]
[0127] In formula (7), u is the input; y is the system output; is the lumped disturbance, and b in equation (7) is replaced by the control gain estimate b0 which is easy to obtain;
[0128] The three second-order systems have the same form;
[0129] Select state variable x1=y, Then the system shown in the above equation can be written as:
[0130]
[0131] As the key core of ADRC, designing an ESO with strong estimation and compensation capability plays a decisive role in the performance of ADRC. Based on formula (8), the third-order linear ESO is constructed as follows:
[0132]
[0133] In formula (9), z1, z2 and z3 are the output y and the output change rate. and the tracking signal of the lumped disturbance f; ξ1, ξ2 and ξ3 are the error correction gains of the ESO;
[0134] The traditional linear SEF is constructed as:
[0135] u0=k p (y-z1)-k d z2 (10)
[0136] In formula (10), k p and k d is the controller parameter;
[0137] The lumped disturbance compensation link is designed as follows:
[0138]
[0139] According to the bandwidth method of linear observer, all poles of ESO are configured to ω c At this point, the following configuration is made for equation (8):
[0140]
[0141] The above configuration simplifies the three adjustable parameters of the observer to one.
[0142] Step 3: IFNLESO design;
[0143] Step 4: Design of CSTSMC feedback control law.
[0144] Example 3
[0145] This embodiment provides a sliding mode anti-disturbance control method for an energy storage converter based on an improved ESO. On the basis of Embodiment 1-2, step 3 is specifically as follows:
[0146] In view of the problems of insufficient accuracy and response speed of traditional linear ESO when facing high-order complex systems, improvements are made on the traditional linear third-order ESO by adding a state variable z4 for the rate of change of the lumped disturbance to observe the changing trend of the lumped disturbance, make effective compensation in advance, improve the observation accuracy, and enhance the anti-disturbance capability; at the same time, in order to increase the observation speed, a nonlinear function is added to the linear observer to improve the convergence speed and accuracy of the ESO. Therefore, a nonlinear fal function is added on the basis of the fourth-order linear ESO to improve the performance of the ESO;
[0147] The rate of change of the lumped disturbance of system (8) is Expanded to x4, the system is reconstructed as follows:
[0148]
[0149] The fourth-order linear ESO equations constructed for equation (13) are:
[0150]
[0151] In formula (14), z1, z2, z3 and z4 are output y, output change rate The aggregate disturbance f and the rate of change of the aggregate disturbance The tracking signal; ξ1, ξ2, ξ3 and ξ4 are the error gains of ESO;
[0152] Assume that in equation (14) is a constant ω F , the fourth-order linear ESO observation error is obtained as:
[0153]
[0154] When the error system enters the steady state, the steady-state error |e1|=ω is obtained. F / ξ4,|e2|=ξ1ω F / ξ4,|e3|=ξ2ω F / ξ4,|e4|=ξ3ω F / ξ4;
[0155] In order to further accelerate the convergence speed of the observer and improve the accuracy, and enhance the observation ability of the observer, an IFNLESO is proposed for the system shown in equation (13):
[0156]
[0157] In formula (16), α1=3 / 4, α2=2 / 4, α2=1 / 4, and the nonlinear function fal is as follows:
[0158]
[0159] In formula (17), δ is the error bound. The fal function has the advantages of “large error, small gain; small error, large gain”. When |e|≤δ, IFNLESO works in the linear domain, which can avoid high-frequency oscillation when the error e is very small. When |e|>δ, the observer works in the nonlinear domain, ensuring that the observer can quickly converge to the error bound.
[0160] The state error equation obtained from equations (13) and (16) is:
[0161]
[0162] And let fal=e a signe(), when the error system enters the steady state, the steady-state errors of the observer are |e1|=(ω F / ξ4) 4 , e2=ξ3(ω F / ξ4) 4 , e3=ξ2(ω F / ξ4) 3 , e4=ξ3(ω F / ξ4) 2 ;when When compared with the observation error of linear ESO, it can be seen that IFNLESO has a larger error magnitude, faster convergence speed and higher accuracy.
[0163] The IFNLESO structure diagram is as follows: Figure 4 shown.
[0164] Example 4
[0165] This embodiment provides a sliding mode self-anti-disturbance control method for an energy storage converter based on an improved ESO. On the basis of Embodiments 1-3, in order to improve the tracking accuracy and robustness of the system, and to improve the anti-disturbance and response speed of the system, a complementary sliding mode control CSMC (Complementary Sliding Mode Control) is introduced. However, due to the switching action of the discontinuous switching function in the traditional sliding mode switching law, jitter is generated, which affects the stability of the system. At the same time, the super-twisting sliding mode control STSMC (Super-Twisting Sliding Mode Control) combines the switching function with an integral link to eliminate high-frequency switching terms, so that the control can suppress jitter well while retaining the advantages of the traditional sliding mode. Therefore, the STSMC law is introduced to replace the switching control law of the traditional CSM, and a complementary super-twisting sliding mode control CSTSMC (Complementary Super-Twisting Sliding Mode Control) is proposed.
[0166] Step 4 is as follows:
[0167] The error state equation is constructed by equations (10) and (12):
[0168]
[0169] Define the generalized sliding surface s g (t) is:
[0170]
[0171] In formula (32), β is the sliding surface constant, and β satisfies >0.
[0172] Design a complementary sliding surface s that is orthogonal to the generalized sliding surface c (t) is:
[0173]
[0174] According to S g and S c The total sliding surface S is obtained m :
[0175] s m =s g +s c =2ε(34)
[0176] Determine g and c The relationship between derivatives:
[0177]
[0178] Choose a Lyapunov function:
[0179]
[0180] Taking the derivative of Lyapunov, we get:
[0181]
[0182] Under the condition that the Lyapunov function satisfies the stability condition, the complementary sliding mode control law is obtained:
[0183] u t =u eq +u rl (38)
[0184] In formula (38), u eq Expressed as the sliding mode equivalent control law:
[0185]
[0186] Taking the super-helical sliding mode control law as the CSTSMC switching control law, then u in equation (33) rl Designed for:
[0187]
[0188] In formula (40), sign(·) is the sign function, and ρ and κ are positive real numbers.
[0189] The control strategy block diagram of CSTSMC is as follows: Figure 5 shown.
[0190] Simulation and experimental verification
[0191] In order to further compare and verify the effectiveness and better control ability of the control method of the present invention, Figure 6 As shown in the figure, a hardware-in-the-loop system based on the MT 6020 real-time simulator was built. The power system model was run in the MT 6020, and the control algorithm was written in the DSPTMS32F28335 controller. Data was transmitted through the physical IO port to achieve closed-loop control, and the output waveform of the power system operation in the simulator was displayed on the Tektronix oscilloscope. The circuit parameters are shown in Table 1, and the circuit control parameters under different control strategies are shown in Table 2.
[0192] Table 1 Circuit topology parameters
[0193]
[0194] Table 2 Control strategy parameters
[0195]
[0196] In Table 2, k p , k i and k d k is the proportional coefficient, integral coefficient and differential coefficient of the bidirectional DC / DC converter. id , k iq K is the integral coefficient of the d-axis and q-axis of the inner current loop of the bidirectional DC / AC converter. pd , k pq , k id , k iq is the proportional coefficient and integral coefficient of the d-axis and q-axis of the inner current loop, k dd , k dq They represent the differential coefficients of the d-axis and q-axis of the inner current loop. In Table 2, α=0.25, δ=0.01 are selected, and the maximum error e max =0.1, disturbance bandwidth ω=130, according to equations (26) to (29), we get λ=300, D0=30.
[0197] (I) Comparative analysis of DC bus voltage fluctuations
[0198] In order to further compare and verify the superiority of the control method of the present invention, the control method of the present invention is used to control PCS with PI control and traditional ADRC, and their transient performances are compared and analyzed.
[0199] Figure 7 shows the DC bus voltage u under the three control strategies when the load power of the PCS changes suddenly between 30KW and 70KW in the inverter state. dc Transient response and PCS AC side a phase AC voltage u a and phase a alternating current i a Transient waveform. The transient process comparison results are as follows Figure 7(a) to Figure 7(c) As shown, it can be seen that the transient adjustment time and overshoot of the traditional ADRC are reduced compared with the PI control, but the control method of the present invention has a shorter adjustment time and a smaller overshoot, showing a better control ability.
[0200] Figure 8 shows the transient response waveforms of the DC bus voltage and the PCS AC side voltage and current under the two control strategies during the switching process between the rectifier and inverter working states. Figures 8(a) to 8(c) It can be seen that the control method of the present invention has a smaller overshoot of the DC bus voltage, a shorter transient time, and a stronger robustness than the traditional ADRC strategy and PI control strategy.
[0201] From the above power mutation experiment comparison, it can be seen that the control method of the present invention is superior to PI control and traditional ADRC in controlling PCS in terms of overshoot and adjustment time. Therefore, the control method of the present invention has better anti-interference performance, more outstanding control ability and stronger robustness.
[0202] (II) Comparative analysis of power response speed
[0203] In order to verify the power response speed under the three control strategies, the power response speed is characterized by the change of d-axis current. Figure 9 shows the d-axis current i when the load power suddenly changes. d and q-axis current i q The transient waveform is shown in FIG9(c). As shown in the power response comparison result, when the load power of the PCS changes suddenly between 30KW and 70KW in the inverter state, the control method of the present invention controls the PCS, and the adjustment time of the d axis is shorter and the overshoot is smaller, that is, the power response speed is faster, and the performance is better than the PI control and the traditional ADRC strategy.
[0204] Figure 10 shows the d-axis current i under three control strategies when switching between the PCS inverter output 30kW state and the rectifier energy storage 30kW state. d and q-axis current iq Compared with PI control and ADRC, the d-axis current transient time under the control method of the present invention is shorter, the overshoot is smaller, and the power response speed is faster.
[0205] As shown in FIG. 11 , under the control method of the present invention, the transient time of the d-axis current is shorter and the overshoot is smaller, that is, under the control method of the present invention, the power response speed of the PCS AC side is faster, which verifies the superiority of the control method of the present invention.
[0206] From the above content, it can be seen that the sliding mode anti-disturbance control method of the energy storage converter based on the improved ESO of the present invention 1) reduces the THD of the output current, shortens the transient time, reduces the harmonics of the system output AC current, and improves the output power quality of the PCS AC side;
[0207] 2) Compared with the PI control strategy and ADRC, the PSC under the control method of the present invention has a smaller overshoot of the DC voltage waveform and a shorter adjustment time when a sudden load change occurs, which can well suppress voltage fluctuations and has good anti-interference performance. At the same time, the control method of the present invention also improves the power response speed and increases the dynamic performance of the system;
[0208] 3) When controlling the PCS to switch between rectification and inversion working states, the impact on the DC bus voltage is reduced, the response speed is improved, the voltage fluctuation is reduced, and the power response switching is faster and smoother.
Claims
1. The sliding mode anti-disturbance control method of energy storage converter based on improved ESO is characterized by: The energy storage converter adopts a dual-loop control combining the voltage outer loop and the current inner loop. The CSTSM-ADRC is designed as the controller of the voltage outer loop and the current inner loop to control the bidirectional AC / DC converter to stabilize the DC bus voltage and realize the energy conversion between the DC side and the AC side. The bidirectional DC / DC converter adopts the PI current single-loop control strategy to control the output power of the energy storage battery through the bidirectional DC / DC converter to match the power required by the load. The design of CSTSM-ADRC is as follows: first, the mathematical model of the converter is fitted as a second-order active disturbance rejection model, and an improved fourth-order nonlinear extended state observer is designed to achieve more accurate estimation and compensation of the lumped disturbance and enhance the anti-disturbance capability. The response speed is improved by using nonlinear functions, and the observer parameters are optimized. Secondly, a complementary super-helical sliding mode feedback control law is designed, and the complementary sliding surface and the super-helical sliding mode algorithm are combined as the switching law to effectively suppress the chattering.
2. The sliding mode anti-disturbance control method for energy storage converter based on improved ESO according to claim 1 is characterized in that: The CSTSM-ADRC design is implemented in the following steps: Step 1: Establish PCS model; Step 2: Simplify the adjustable parameters of the observer based on the ADRC principle; Step 3: IFNLESO design; Step 4: Design of CSTSMC feedback control law.
3. The sliding mode anti-disturbance control method of energy storage converter based on improved ESO according to claim 2 is characterized in that: The step 1 specifically comprises: Under three-phase voltage balance, the PCS topology establishes a mathematical model in a three-phase stationary coordinate system described by the switching function: In formula (1), S j is the switching function, defined as: The PCS mathematical model in the d and q two-phase rotating coordinate system is obtained through Park transformation: In formula (2), v d 、v q Respectively represent the components of the AC grid voltage on the d and q axes; i d 、i q are the components of the AC grid-side current on the d and q axes respectively; S d , S q are the electrical switching functions on the d and q axes on the grid side respectively; ω is the grid voltage angular frequency.
4. The sliding mode anti-disturbance control method for energy storage converter based on improved ESO according to claim 3 is characterized in that: The step 2 is specifically as follows: In order to achieve the unity power factor of the unit and connect to the grid, the reactive current is set to zero. From equation (3), we know that the mathematical model of the bidirectional AC / DC converter is a second-order system. We take the derivative of equation (3) and fit the voltage outer loop and the d-axis and q-axis current inner loops as second-order systems containing disturbances. The state variables of the voltage loop system are defined as Current loop system state variables The obtained voltage loop second-order system is: The second-order current loop system is: In formula (4) and formula (5), b d and b q are the control gains of the d-axis and q-axis respectively; the internal influence F vd , external disturbance F d and the lumped disturbance F of various unknown inaccurate models q for: For the second-order system design shown in Equation (4) and Equation (5), the traditional linear ADRC is designed as a second-order system including disturbances. The second-order system is as follows: In formula (7), u is the input; y is the system output; is the lumped disturbance, and b in equation (7) is replaced by the control gain estimate b0 which is easy to obtain; The three second-order systems have the same form; Select state variables Then the system shown in the above equation can be written as: Based on formula (8), the third-order linear ESO is constructed as follows: In formula (9), z1, z2 and z3 are the output y and the output change rate. and the tracking signal of the lumped disturbance f; ξ1, ξ2 and ξ3 are the error correction gains of the ESO; The traditional linear SEF is constructed as: u0=k p (y-z1)-k d z2 (10) In formula (10), k p and k d is the controller parameter; The lumped disturbance compensation link is designed as follows: According to the bandwidth method of linear observer, all poles of ESO are configured to ω c At this point, the following configuration is made for equation (8): The above configuration simplifies the three adjustable parameters of the observer to one.
5. The sliding mode anti-disturbance control method for energy storage converter based on improved ESO according to claim 4 is characterized in that: The step 3 is specifically as follows: Improvements are made to the traditional linear third-order ESO by adding a state variable z4, which is the rate of change of the lumped disturbance, to observe the changing trend of the lumped disturbance, make effective compensation in advance, improve the observation accuracy, and enhance the anti-disturbance capability. At the same time, in order to increase the observation speed, a nonlinear function is added to the linear observer to improve the convergence speed and accuracy of the ESO. Therefore, a nonlinear fal function is added on the basis of the fourth-order linear ESO to improve the performance of the ESO. The rate of change of the lumped disturbance of system (8) is Expanded to x4, the system is reconstructed as follows: The fourth-order linear ESO equations constructed for equation (13) are: In formula (14), z1, z2, z3 and z4 are output y, output change rate The aggregate disturbance f and the rate of change of the aggregate disturbance The tracking signal; ξ1, ξ2, ξ3 and ξ4 are the error gains of ESO; Assume that in equation (14) is a constant ω F , the fourth-order linear ESO observation error is obtained as: When the error system enters the steady state, the steady-state error |e1|=ω is obtained. F / ξ4,|e2|=ξ1ω F / ξ4,|e3|=ξ2ω F / ξ4,|e4|=ξ3ω F / ξ4; In order to further accelerate the convergence speed of the observer and improve the accuracy, as well as the observation ability of the observer, an IFNLESO is proposed for the system shown in equation (13): In formula (16), α1=3 / 4, α2=2 / 4, α2=1 / 4, and the nonlinear function fal is as follows: In formula (17), δ is the error bound; The state error equation obtained from equations (13) and (16) is: And let fal=e a signe(), when the error system enters the steady state, the steady-state errors of the observer are |e1|=(ω F / ξ4) 4 , e2=ξ3(ω F / ξ4) 4 , e3=ξ2(ω F / ξ4) 3 , e4=ξ3(ω F / ξ4) 2 .
6. The sliding mode anti-disturbance control method for energy storage converter based on improved ESO according to claim 5 is characterized in that: The step 4 is specifically as follows: The error state equation is constructed by equations (10) and (12): Define the generalized sliding surface s g (t) is: In formula (32), β is the sliding surface constant, and β satisfies > 0; Design a complementary sliding surface s that is orthogonal to the generalized sliding surface c (t) is: According to s g and c Get the total sliding surface s m : s m =s g +s c =2ε(34) Determine g and c The relationship between derivatives: Choose a Lyapunov function: Taking the derivative of Lyapunov, we get: Under the condition that the Lyapunov function satisfies the stability condition, the complementary sliding mode control law is obtained: in t =in eq +in rl (38) In formula (38), u eq Expressed as the sliding mode equivalent control law: Taking the super-helical sliding mode control law as the CSTSMC switching control law, then u in equation (33) rl Designed for: In formula (40), sign(·) is the sign function, and ρ and κ are positive real numbers.
