Distributed self-adaptive fault-tolerant control method for direct-current micro-grid

Abstract

The invention discloses a distributed adaptive fault-tolerant control method for a direct-current micro-grid, and the method comprises the steps: designing a distributed adaptive fault compensator based on an extended state observer, and enabling the distributed adaptive fault compensator to be used for compensating the control input of each distributed power supply under the conditions of actuator failure, bias fault and external interference; based on a distributed consistency protocol and a self-adaptive fault compensator, a distributed secondary comprehensive performance optimization adjustment mechanism is provided, and the fault-tolerant capability, the voltage quality and the current sharing precision of the direct-current micro-grid system under the actuator fault are improved; and a Lyapunov stability proof of system convergence based on a distributed adaptive fault-tolerant control method is given. According to the method, multiple types of actuator faults are considered, so that the method is more general; upper bound or lower bound information of any fault is not needed, the difficulty of parameter configuration of the controller can be effectively reduced, and the method is more universal and practical; the distributed power supply is completely distributed, no global information is needed, only the information of the local and adjacent distributed power supplies is utilized, and the communication cost is reduced.

Description

Technical Field

[0001] The present invention belongs to the technical field of fault-tolerant control of direct current microgrids, and in particular relates to a distributed adaptive fault-tolerant control method for direct current microgrids. Background Art

[0002] In recent years, due to the DC power generation characteristics of renewable energy and the popularity of DC loads, DC microgrids have achieved rapid development. Isolated DC microgrids are considered to be the most effective way to solve the power supply problem in remote mountainous areas or islands. In the case of dual uncertainty of distributed renewable energy output and load demand without the support of the main power grid, compared with the high communication dependence and computational complexity of centralized control, distributed control relying on sparse communication networks has good scalability and flexibility, and can better solve the problems of voltage drop, current balancing, and fault tolerance in DC microgrids. On the other hand, the actuators in the DC microgrid, the key equipment for the system to realize energy conversion, will inevitably fail, which may lead to the inability to achieve the control target, and even endanger the stable operation and safety and reliability of the DC microgrid. Therefore, it is crucial to establish an effective fault-tolerant mechanism.

[0003] The existing fault-tolerant control methods of DC microgrids mainly include sliding mode control, H∞ control and adaptive control. The sliding mode control method controls the movement of the system state by introducing a specific sliding surface, which can achieve fast sliding and accurate tracking of the system state, but there are problems such as jitter and complex control parameter design, which can easily bring additional noise and oscillation in practical applications. The H∞ control method can combine the robust stability and control performance of the system in a unified design framework, avoiding repeated optimization, but this method focuses on minimizing the gain in the worst case and has strong conservatism. The adaptive control method does not require accurate fault information, and can automatically adjust the control input only based on the system state change caused by the fault; it can also use observer technology to estimate the fault online, and then offset the impact of the fault on the system by designing compensation items. The design is simple and can optimize the control performance. In addition, most of the existing microgrid actuator fault-tolerant control methods are only for actuator bias faults, and do not consider the impact of actuator failure faults.

[0004] The patents that have been applied for are as follows:

[0005] Patent number: CN 114447909 A, Patent name A DC microgrid distributed fault-tolerant control method and controller based on a fault observer, used for DC microgrids to achieve fault-tolerant control of DC bus voltage tracking and current proportional distribution in the event of an actuator failure. This method does not require the communication resources of the control system, can complete the real-time estimation of the actuator fault signal and correct the control law, effectively resist the impact of actuator faults on voltage tracking and current distribution, and improve the fault tolerance performance and safety margin of the DC microgrid.

[0006] The controller design of this method is complex, the parameter setting is difficult, and the amplitude of the first-order derivative of the fault signal is required to be bounded, which is a strict assumption within a limited time. In addition, although the use of distributed fault-tolerant control does not rely on complex communications, it is difficult to achieve zero-error control and make global optimal decisions due to the lack of communication information between devices and the global operating status.

[0007] The distributed adaptive fault-tolerant control method proposed in the present invention uses adaptive technology to estimate faults online and compensate in real time. It is simple in design and can optimize control performance. It only requires the boundedness of the fault and does not require any upper or lower bound information of the fault. It can effectively reduce the difficulty of controller parameter configuration and is more versatile and practical. The distributed control adopted only uses the information of local and neighboring distributed power sources, which reduces the communication cost, has good scalability and flexibility, and can take into account the system's voltage quality and current sharing and even global optimization, thereby improving the robustness of the DC microgrid system under actuator failure.

[0008] Patent number: CN118399476A, patent name: Distributed fault-tolerant secondary control method for isolated microgrid based on event triggering. The method uses a model that takes into account both partial actuator failure and bias fault droop control, and studies the fault-tolerant control of distributed AC microgrids, so as to achieve the system's restoration of normal frequency and voltage in the event of an actuator failure, and to achieve accurate power distribution according to the capacity of each power supply in the system. The method adopts a distributed fault-tolerant secondary control strategy based on event triggering, establishes the triggering conditions for event triggering, and exchanges information between each distributed power supply at the time of event triggering, reducing the communication pressure between distributed power supplies.

[0009] This method studies the fault-tolerant control of distributed AC microgrids. The definition of events and the setting of trigger conditions require careful design and debugging. The event triggering control mechanism of distributed systems is more complex, which brings difficulties to the design of event triggering conditions.

[0010] The fault-tolerant control method proposed in the present invention is aimed at the fault-tolerant control of distributed DC microgrids. It can realize online estimation and real-time compensation of actuator faults, and improve the fault tolerance, voltage quality and current sharing accuracy of the DC microgrid system under actuator failure, bias fault and external interference.

[0011] Patent number: CN112346439B, patent name: Microgrid sensor fault-tolerant control method based on PD-type learning observer. According to the state space model of the microgrid inverter distributed power supply, a system closed-loop output feedback controller is established; a microgrid system sensor fault model is established, and the sensor fault model is converted into an actuator fault model form; and a sensor fault-tolerant PD-type learning observer is established to obtain the reconstructed fault signal, and the output after reconstructed signal compensation is fed back to the output feedback controller to ensure stable tracking of the output power.

[0012] This method essentially uses a PD-type learning observer to monitor sensor faults converted into actuator faults in real time, and uses the reconstructed fault signal to correct the sensor's measurement error, which increases the complexity of system design.

[0013] The adaptive fault-tolerant control method proposed in the present invention directly uses adaptive technology to perform online estimation and real-time compensation for faults. It is simple in design and can optimize control performance. It only requires the boundedness of the fault and does not require any upper or lower bound information of the fault. It can effectively reduce the difficulty of controller parameter configuration and is more universal and practical. Summary of the invention

[0014] In order to solve the above problems, the present invention proposes a distributed adaptive fault-tolerant control method for a DC microgrid. The method takes into account multiple types of actuator failures, including partial actuator failure, complete failure, bias failure, stuck failure and external disturbance, and is more general. It is also fully distributed and does not require any global information. It only uses information from local and neighboring distributed power sources, reducing communication costs.

[0015] To achieve the above object, the technical solution adopted by the present invention is:

[0016] A distributed adaptive fault-tolerant control method for a DC microgrid comprises the following steps:

[0017] Step 1) Based on standard distributed quadratic control, the quadratic control problem of the DC microgrid is transformed into a consistency problem of the first-order linear MAS system;

[0018] The analysis steps of consistency problem are as follows:

[0019] First, the distributed secondary control is introduced into the traditional droop control for compensation to eliminate the deviation caused by the droop control:

[0020] V i ref =V rated -m i I oi +δ i (1)

[0021] Where V i ref The voltage reference value generated by the droop control, V rated is the rated value of the DC bus voltage, m i is the droop coefficient, I oi For DG i Output current, δ i =∫u i dt is the compensation term of distributed quadratic control, u i is the control input;

[0022] Then, the difference between the average output voltage and the rated voltage and the current consistency error between the DGs are used as the compensation control input for the standard distributed secondary control:

[0023]

[0024] In the formula, if DG i Can directly receive the reference signal, then b i >0 holds, otherwise b i =0;c i >0 is the gain coefficient; is the average output voltage of each DG, observed by the voltage observer (3) based on the average consistency protocol; It's DG i The derivative of the output voltage;

[0025] Usually the bandwidth of the voltage loop is much larger than the bandwidth of the droop control loop, which makes the output voltage V i Fast tracking droop control generates a reference voltage value V i ref , that is, V i =V i ref Therefore, by performing simple calculations on equations (1), (2), and (3), we can obtain:

[0026]

[0027] Finally, in the steady state, m i I oi It will converge to a constant value kI ss , so define and θ ref =kIss , then the consistency problem of DC microgrid can be expressed as:

[0028]

[0029] definition is a distributed consistency protocol, then

[0030] Step 2) Based on the extended state observer, a distributed adaptive fault compensator is designed for the control input of each distributed power source under actuator failure, bias fault and external interference;

[0031] Step 3) Based on the distributed consistency protocol and adaptive fault compensator, a distributed secondary comprehensive performance optimization adjustment mechanism is proposed to improve the robustness, voltage quality and current sharing accuracy of the DC microgrid under actuator faults.

[0032] As a further improvement of the present invention, the specific steps of step 2) are as follows:

[0033] First, consider the erroneous control input under actuator failure as:

[0034]

[0035] In the formula, represents the erroneous control input under actuator failure; ρ i is such that 0≤ρ i The unknown coefficient ≤1 indicates the failure of the i-th actuator; η i Represents bias unknown time-varying faults, expressed as step, ramp, sine and their combination;

[0036] Then, establish the extended state observer:

[0037]

[0038] In the formula, Indicates DG i The observed value of the state variable, λ represents the observer gain to be determined, ξ i represents the relative output estimation error of the i-th observer;

[0039] Finally, based on the relative output estimation error of the extended state observer, a distributed adaptive fault compensator is designed for actuator failure, bias fault and external interference:

[0040]

[0041] In the formula, and represents the adaptive learning rate, ι i To satisfy The non-negative time-varying scalar of is, and the corresponding adaptive learning rate is expressed as:

[0042]

[0043] In the formula, γ i1 and γ i2 is the normal number to be designed.

[0044] As a further improvement of the present invention, the specific steps of step 3) are as follows:

[0045] First, based on the distributed consistency protocol and adaptive fault compensator, a distributed secondary comprehensive performance optimization adjustment mechanism is proposed:

[0046]

[0047] Combining equations (5), (6) and (10), the dynamic consistency protocol under actuator failure can be expressed as:

[0048]

[0049] Then, define the global state error e = col{e1,e2,…,e N}, where e i =θ i -θ ref Indicates DG i The state tracking error of , can be obtained from formula (11) i The derivative of is:

[0050]

[0051] Then the global state error is:

[0052]

[0053] Finally, the Lyapunov function is constructed to prove that the proposed distributed quadratic comprehensive performance optimization adjustment mechanism can make the global state error e converge to zero asymptotically.

[0054]

[0055] In the formula, μ i1 ,μ i2 ,k i1 and k i2 is a positive constant to be determined.

[0056] Beneficial effects:

[0057] A distributed adaptive fault-tolerant control method for a DC microgrid proposed in the present invention takes into account multiple types of actuator failures, including partial actuator failure, complete failure, bias failure, stuck failure and external disturbance, and is more general; and does not require any upper or lower bound information of the fault, which can effectively reduce the difficulty of controller parameter configuration, and is more universal and practical; it is also fully distributed, does not require any global information, and only uses information from local and neighboring distributed power sources, which reduces communication costs and improves the fault tolerance, voltage quality and current sharing accuracy of the DC microgrid system under actuator failures. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 is a flow chart of the present invention;

[0059] Figure 2 This is a control block diagram of a single distributed power supply in a DC microgrid of the present invention;

[0060] Figure 3 It is a schematic diagram of the structure of a ring-shaped DC microgrid including four distributed power sources of the present invention;

[0061] Figure 4 The output current and output voltage response curves of the present invention under various types of actuator faults;

[0062] Figure 5 The output current and output voltage response curves of the present invention under load changes and actuator failures are shown. DETAILED DESCRIPTION

[0063] The present invention is further described in detail below with reference to the accompanying drawings and specific embodiments. Figure 1 It is the flow chart of the present invention:

[0064] Embodiment 1:

[0065] This embodiment and the following embodiments are all based on Figure 2 The DC microgrid shown is used as an object for explanation. Figure 3 The DC microgrid shown is a ring DC microgrid containing four distributed power sources with a voltage level of 48V. Each distributed power source is connected to the bus through a DC / DC converter. Figure 3 As shown, DG1-DG4 are followers, the leader is a virtual micro-source, and only DG1 can obtain the reference information of the leader.

[0066] against Figure 3 The system shown in the figure considers various types of actuator failures to evaluate the effectiveness of the fault-tolerant control method of the present invention. The droop coefficient of each distributed power source is set to m1=m2=m3=m4=0.5, then the output current ratio of each DG should be 1:1:1:1. The simulation scenario is set as follows:

[0067] 1) At t = 1s, the actuator of DG1 fails u1 F =0.5u1+3t;

[0068] 2) At t = 3s, the actuator of DG4 fails u4 F =u4-10;

[0069] 3) At t = 5s, the actuator of DG3 fails u3 F =u3+10sin(3πt)+5.

[0070] Figure 4 The response curves of the DC microgrid system under various types of actuator faults are shown. Figure 4 It can be seen that during t = 0 ~ 1s, the output current of each DG can quickly converge to the steady-state value, the output voltage remains stable, and the average voltage is adjusted to the rated voltage value. After t = 1s, as the types and number of actuator faults increase, the system can still quickly tolerate various types of actuator faults and maintain stable operation. The output voltage and current waveforms do not fluctuate significantly, achieving the expected goals of current balancing, voltage regulation, and fault tolerance.

[0071] Embodiment 2:

[0072] This embodiment considers load changes and actuator failures to evaluate the performance of the DC microgrid system using the fault-tolerant control method of the present invention. The droop coefficient of each distributed power source is set to m1=2m2=m3=2m4=1, then the output current ratio of each DG should be 1:2:1:2. The simulation scenario is set as follows:

[0073] 1) At t = 1s, the actuator of DG1 fails u1 F =0.3u1+20;

[0074] 2) At t = 2s, the load of DG1 increases from 15Ω to 30Ω;

[0075] 3) At t = 3s, the actuator of DG4 fails u4 F =0.8u4+10sin(2πt+π)-15;

[0076] 4) At t=4s, the load of DG3 decreases from 20Ω to 10Ω.

[0077] The system response curve is as follows Figure 5As shown in the figure, before t = 1s, the output current and voltage of each DG have reached a steady state, and the output current ratio is 1:2:1:2. At t = 2s, as the local load of DG1 increases, the output current of each DG decreases and then reaches a new steady state. Similarly, at t = 4s, as the local load of DG3 decreases, the output current of each DG increases and reaches a new steady state. In addition, it can be seen that the actuator fault has almost no effect on the system output, and the expected control target is achieved.

[0078] The above description is only a preferred embodiment of the present invention and does not constitute any other form of limitation to the present invention. Any modification or equivalent change made based on the technical essence of the present invention still falls within the scope of protection required by the present invention.

Claims

1. A distributed adaptive fault-tolerant control method for a DC microgrid, characterized in that: The following steps are involved: Step 1) Based on standard distributed quadratic control, the quadratic control problem of the DC microgrid is transformed into a consistency problem of the first-order linear MAS system; The analysis steps of consistency problem are as follows: First, the distributed secondary control is introduced into the traditional droop control for compensation to eliminate the deviation caused by the droop control: In the formula, The voltage reference value generated by the droop control, V rated is the rated value of the DC bus voltage, m i is the droop coefficient, I oi For DG i Output current, δ i =∫u i dt is the compensation term of distributed quadratic control, u i is the control input; Then, the difference between the average output voltage and the rated voltage and the current consistency error between the DGs are used as the compensation control input for the standard distributed secondary control: In the formula, if DG i Can directly receive the reference signal, then b i >0 holds, otherwise b i =0;c i >0 is the gain coefficient; is the average output voltage of each DG, observed by the voltage observer (3) based on the average consistency protocol; It's DG i The derivative of the output voltage; Usually the bandwidth of the voltage loop is much larger than the bandwidth of the droop control loop, which makes the output voltage V i Fast tracking of the reference voltage value generated by droop control Right now Therefore, by performing simple calculations on equations (1), (2) and (3), we can obtain: Finally, in the steady state, m i I oi It will converge to a constant value kI ss , so define and θ ref =kI ss , then the consistency problem of DC microgrid can be expressed as: definition is a distributed consensus protocol, then Step 2) Based on the extended state observer, a distributed adaptive fault compensator is designed for the control input of each distributed power source under actuator failure, bias fault and external interference; Step 3) Based on the distributed consistency protocol and adaptive fault compensator, a distributed secondary comprehensive performance optimization adjustment mechanism is proposed to improve the robustness, voltage quality and current sharing accuracy of the DC microgrid under actuator faults.

2. A DC microgrid distributed adaptive fault-tolerant control method according to claim 1, characterized in that: Step 2) The specific steps are as follows: First, consider the erroneous control input under actuator failure as: In the formula, represents the erroneous control input under actuator failure; ρ i is such that 0≤ρ i The unknown coefficient ≤1 indicates the failure of the i-th actuator; η i Represents bias unknown time-varying faults, expressed as step, ramp, sine and their combination; Then, establish the extended state observer: In the formula, Indicates DG i The observed value of the state variable, λ represents the observer gain to be determined, ξ i represents the relative output estimation error of the i-th observer; Finally, based on the relative output estimation error of the extended state observer, a distributed adaptive fault compensator is designed for actuator failure, bias fault and external interference: In the formula, and represents the adaptive learning rate, ι i To satisfy The non-negative time-varying scalar of is, and the corresponding adaptive learning rate is expressed as: In the formula, γ i1 and γ i2 is the normal number to be designed.

3. A DC microgrid distributed adaptive fault-tolerant control method according to claim 1, characterized in that: Step 3) The specific steps are as follows: First, based on the distributed consistency protocol and adaptive fault compensator, a distributed secondary comprehensive performance optimization adjustment mechanism is proposed: Combining equations (5), (6) and (10), the dynamic consistency protocol under actuator failure can be expressed as: Then, define the global state error e = col{e1,e2,…,e N }, where e i =θ i -θ ref represents the state tracking error of DGi. From formula (11), we can get e i The derivative of is: Then the global state error is: Finally, the Lyapunov function is constructed to prove that the proposed distributed quadratic comprehensive performance optimization adjustment mechanism can make the global state error e converge to zero asymptotically. In the formula, μ i1 ,μ i2 ,k i1 and k i2 is a positive constant to be determined.

Images