CN-122017452-A - Anti-outlier memory type event-triggered fault detection method for photovoltaic micro-grid

CN122017452ACN 122017452 ACN122017452 ACN 122017452ACN-122017452-A

Abstract

The invention discloses a fault detection method of a photovoltaic micro-grid based on anti-outlier memory type event triggering. The method comprises the steps of 1, establishing a state space model of a photovoltaic micro-grid single-phase inverter with external disturbance and fault signals, 2, designing an outlier-resistant memory type event trigger communication mechanism based on historical data mean values aiming at noise pollution existing in measurement signals and outlier caused by environmental mutation, reducing excessive triggering caused by measurement noise, avoiding false triggering and false alarm caused by the outlier, 3, establishing a photovoltaic micro-grid closed-loop fault detection system model based on a residual filter, solving parameters of the residual filter and the event trigger mechanism by means of a Liapunov method, a Simpson rule and a linear matrix inequality technology, and adopting the designed outlier-resistant memory type event trigger mechanism to carry out fault detection of the photovoltaic micro-grid. Compared with the existing event triggering mechanism, the invention provides an anti-outlier memory type event triggering mechanism, which effectively reduces the triggering and updating frequency of detection signals, improves the fault detection precision and has certain engineering application value.

Inventors

YAN SHEN
ZHANG KUN

Assignees

南京林业大学
南京众行能源科技有限公司

Dates

Publication Date: 20260512
Application Date: 20260114

Claims (2)

1. The fault detection method of the photovoltaic micro-grid based on anti-outlier memory type event triggering is characterized by comprising the following steps of: S1, establishing a state space model of a single-phase inverter of a photovoltaic micro-grid; s2, designing a memory type event triggering communication mechanism of an anti-outlier based on a historical data mean value; and S3, deriving a closed-loop residual filtering system model, solving parameters of a residual filter and an event triggering mechanism, and detecting faults of the photovoltaic micro-grid. The fault detection method of the photovoltaic micro-grid based on the anti-outlier memory type event trigger is characterized in that a state variable of the single-phase inverter can be used as a state variable of the photovoltaic micro-grid. According to kirchhoff voltage and current law, a state space model of the photovoltaic micro-grid can be obtained as follows: Wherein the method comprises the steps of C=[0 1 0]; Y (t) represents a state vector, a control input and a measurement output respectively, i f (t)、R f 、L f represents a current, a resistance and an inductance of an inductance-capacitance filter, R o 、L o represents a resistance and an inductance of a transmission line, i o (t)、v o (t) represents an output current and a voltage, V c (t) represents a capacitance voltage, D represents a duty ratio of a Boost circuit, and V sun represents a storage voltage of a photovoltaic micro grid; Gain for the state controller. Giving state feedback controller gain So that Is a helvetz matrix. Considering the faults g (t) caused by external interference w (t), component aging and the like, the following stabilized closed-loop photovoltaic micro-grid state space model can be obtained: Wherein the method comprises the steps of E. G is a system parameter matrix. The method comprises the steps of adopting the average value of historical data as an input signal of an event trigger mechanism, effectively reducing high-frequency trigger caused by measurement noise, simultaneously introducing the upper bound of an event trigger error term to eliminate false trigger caused by outliers, and improving the accuracy of fault detection, and specifically comprises the following steps: The following memory-type event trigger mechanism against outliers is constructed: wherein: t k h,t k+1 h represents the current and next trigger times, h represents the sampling period, θ represents the interval length of the history data, Respectively representing the historical data average value of the current sampling time and the historical data average value of the triggering time, and the sign Representation of For event trigger error term, σ 1 ,σ 2 represents a trigger threshold parameter. When the event triggering condition (3) is met, the current time mean value signal is triggered and the fault detection filter input is updated to Otherwise, the mean value signal at the current moment is not triggered, and the fault detection filter inputs and maintains the value at the last trigger moment When the interval length of the historical data tends to 0, namely theta-0, and the limiting condition of the upper boundary of the event trigger error term is not considered, the designed trigger mechanism is simplified into the following conventional event trigger mechanism based on the instantaneous measurement signal: Wherein the method comprises the steps of
2. The method for detecting the faults of the photovoltaic micro-grid based on the anti-outlier memory type event triggering according to claim 1 is characterized in that a closed-loop residual error filtering system model is derived, and residual error filter and event triggering mechanism parameters are solved by means of a Leidepro method, a Simpson rule, a matrix transformation technology, a linear matrix inequality and the like, so that a fault detection algorithm is designed to detect and early warn the faults of the photovoltaic micro-grid. The method comprises the following steps: 1) Giving a state space model of the residual filter: Wherein the method comprises the steps of Τ k is the network transmission delay and a d ,B d ,C d ,D d is the filter parameters to be designed. 2) A reference model of the fault signal is introduced to produce a desired residual evaluation result. Fault signals g (t) to Is of the transfer function of Its state space model can be expressed as: 3) Definition of the definition Can be expressed as: 4) The integral term in equation (7) is approximated by the simpson's law, Can be further expressed as: Wherein the method comprises the steps of 5) Definition of the definition The following fault detection closed loop system is available: Wherein the method comprises the steps of 6) In order to solve the parameters of the filter and the event trigger, the following Leidefenov function is selected first: Wherein: 7) The following new vectors are defined: 8) According to the above definition, the closed loop system (9) can be expressed as: 9) The derivative of the Liapunov function is calculated as: 10 Using Jensen inequality and the mutual convex technique, the integral term in equation (13) can be scaled as follows: 11 To have the system meet the H ∞ stability performance, the derivative of the li-apunov function needs to meet: 12 From the event triggering condition (3), it is possible to obtain: Wherein the method comprises the steps of Is a positive constant. 13 In combination with the above-described condition, the establishment of the following matrix inequality form can ensure that (15) is established: 14 By using And constructing a matrix Equation (17) is equivalent to: Wherein: 15 By selecting a matrix The method can be as follows: 16 Definition of (d) And bringing formula (19) into formula (18), it is possible to obtain: Wherein: 17 Finally, the filter parameters can be determined as follows: 18 To design the fault detection algorithm, a residual evaluation function is introduced 19 To detect a fault signal, a fault alarm threshold value needs to be given in advance: 20 Based on the residual evaluation function and the fault alarm threshold, giving the following fault alarm conditions: 21 The event trigger matrix omega and the filter parameters obtained in the formula (21) are brought into a closed loop fault detection system, the fault alarm condition of the design is monitored, and when a residual error evaluation function is adopted Greater than the alarm threshold And generating an alarm signal to prompt the occurrence of faults, otherwise, indicating that the system is in a normal running state.

Description

Anti-outlier memory type event-triggered fault detection method for photovoltaic micro-grid Technical Field The invention relates to a fault detection method of a photovoltaic micro-grid based on anti-outlier memory type event triggering, in particular to an event triggering fault detection method of a networked photovoltaic micro-grid system for measuring signals with outliers. Background With the continuous progress of industrialization, the demand of energy for human society is increasing. Due to the increasing exhaustion of non-renewable resources such as petroleum, natural gas and the like and the international consensus of carbon neutralization, the development and utilization of renewable clean energy sources such as wind energy, solar energy and the like can alleviate energy crisis. However, as the equipment such as the photovoltaic panel is exposed outdoors for a long time, the equipment is easy to suffer from failure problems caused by ageing of components, damage of the components and the like, so that the system cannot safely and stably operate, and the equipment is powered off. Therefore, it is important to study the problem of fault detection of the photovoltaic micro grid system based on the communication network. Meanwhile, an event-triggered communication mechanism with the characteristic of transmission on demand is gradually applied to control and estimation design of a networked micro-grid system, so that transmission and calculation of redundant signals can be effectively reduced, and the resource utilization rate of a communication network is improved. However, when the measurement data is interfered by random fluctuation, the existing event triggering mechanism based on instantaneous data is easy to excessively trigger the measurement signal, and wastes limited communication resources. In addition, when the measurement data has an outlier, false triggering of the measurement signal may be caused, thereby generating false fault alarm information. In order to solve the above problems, it is necessary to re-design a new event trigger transmission mechanism on the premise of ensuring a certain system performance, so as to effectively avoid excessive triggering and false triggering of the measurement signal and improve the accuracy of fault detection. Disclosure of Invention Based on the analysis, the invention constructs a fault detection method of the photovoltaic micro-grid based on anti-outlier memory type event triggering in order to improve the utilization rate of network resources and reduce the false alarm rate of fault detection. The specific technical scheme of the invention comprises the following steps: establishing a linear state space model of the photovoltaic micro-grid system; Designing a memory event trigger mechanism of anti-outliers; And solving parameters of an event trigger and a filter which enable the residual filtering system to be stable, and performing fault detection on the photovoltaic micro-grid system. The fault detection method of the photovoltaic micro-grid based on the anti-outlier memory type event trigger is characterized in that the established state space model is as follows: Wherein the method comprises the steps of Y (t) represents a state vector, a control input and a measurement output respectively, i f(t)、Rf、Lf represents a current, a resistance and an inductance of an inductance-capacitance filter, R o、Lo represents a resistance and an inductance of a transmission line, i o(t)、vo (t) represents an output current and a voltage, V c (t) represents a capacitance voltage, D represents a duty ratio of a Boost circuit, and V sun represents a storage voltage of a photovoltaic micro grid; Gain for the state controller. Selecting state feedback controller gainSo thatIs a helvetz matrix. Considering the faults g (t) caused by external interference w (t), component aging and the like, the following stabilized closed-loop photovoltaic micro-grid state space model can be obtained: Wherein the method comprises the steps of E. G is a system parameter matrix. The fault detection method of the photovoltaic micro-grid based on the anti-outlier memory type event trigger is characterized in that a designed anti-outlier memory type event trigger mechanism is as follows: wherein: t kh,tk+1 h represents the current and next trigger times, h represents the sampling period, θ represents the interval length of the history data, Respectively representing the historical data average value of the current sampling time and the historical data average value of the triggering time, and the signRepresentation ofFor event trigger error term, σ 1,σ2 represents a trigger threshold parameter. When the event triggering condition (3) is met, the current time mean value signal is triggered and the fault detection filter input is updated toOtherwise, the mean value signal at the current moment is not triggered, and the fault detection filter inputs and maintains the value at the last tri