CN116432978A

CN116432978A - Method for calculating power supply reliability index of highway self-consistent energy system

Info

Publication number: CN116432978A
Application number: CN202310469510.2A
Authority: CN
Inventors: 黄仙; 冯璋洁; 纪文童; 叶笑容
Original assignee: North China Electric Power University
Current assignee: North China Electric Power University
Priority date: 2023-04-27
Filing date: 2023-04-27
Publication date: 2023-07-14

Abstract

A method of calculating a power supply reliability index for a highway self-consistent energy system, comprising: establishing a highway self-consistent energy system; constructing a real-time power model of a supply side and a traffic load model of a demand side of a highway self-consistent energy system: constructing a power generation power model of the wind turbine by using a power-wind speed conversion formula of the wind turbine, constructing a power generation power model of the photovoltaic turbine by using a power conversion formula, and constructing a traffic load model by adopting a scene of normal distribution construction; establishing a reliability parameter model of each component part of the expressway self-consistent energy system; and calculating the power supply reliability index by using the three-state sequential Monte Carlo. The three-state sequential Monte Carlo is adopted, so that the three-state sequential Monte Carlo is considered from an application object, the distinction between the power supply element and other element types is reflected from an algorithm, and the limitation that the traditional sequential Monte Carlo algorithm does not distinguish the power supply fault from other element faults such as a switching element is improved by considering various fault types.

Description

Method for calculating power supply reliability index of highway self-consistent energy system

Technical Field

The invention relates to a calculation method, in particular to a calculation method of a power supply reliability index, a calculation method of a power supply reliability index of a highway self-consistent energy system in a planning stage and application thereof.

Background

The conventional sequential monte carlo algorithm is usually fixed history data on the supply side and the demand side of the study object in the process of calculating the power supply reliability index of the power system, and uncertainty of both supply and demand ends is not considered.

Meanwhile, aiming at the algorithm, in the calculation flow of the traditional sequential Monte Carlo algorithm, elements only have two states: and the normal operation and the faults are carried out, and the relation between the power supply and the load is only analyzed in the two periods of normal operation time T1 and fault repair time T2, so that the power supply and other switching elements can only stop the faults when faults occur, and the working capacity of the power supply and other switching elements is instantaneously lost. Such as: authors Guo, liu Sai consider the two states of the element in "sequential Monte Carlo simulation based multisource direct current distribution network reliability analysis" published in the distribution technical journal, and when a fault occurs, the element is stopped instantaneously no matter what fault occurs.

Thus, the conventional sequential Monte Carlo algorithm does not consider the failure mode diversity of the failed component, and considers the failure state of the power supply to be consistent with that of the other components. However, in practice, after most of common fault conditions occur in wind power plants and photovoltaic power plants, a period of time of operation with faults exists, and when some small faults occur in a wind-light generator set, the wind-light generator set does not appear as a shutdown fault, and instantaneously loses the power generation capacity, but appears as weakening the power generation capacity. In the period, in order to ensure the economical efficiency of operation, a crew can analyze the generation reason of the fault under the condition of no shutdown, and shutdown maintenance is performed after the fault problem is found.

According to the eighteenth rule of a wind driven generator overhaul management method (A version) issued by new energy development limited company of China Hua Neng group, only main auxiliary equipment and auxiliary equipment are in failure, and if the failure appears as follows: the wind wheel rotates to generate abnormal sound, the direction of the wind wheel is not flexible or the wind wheel can not turn, the wind wheel is stumbled and slowed down, the wind wheel rotates, and the generator does not generate electricity. However, if the wind turbine generator is in a state that the wind speed reaches the rated wind speed, but the wind wheel does not reach the rated rotation speed and the generator cannot output the rated voltage, the power generation capacity of the wind turbine generator is reduced, the wind turbine generator can be checked and removed, and shutdown processing is not needed immediately.

Also, faults of the photovoltaic unit are classified into two types according to national energy information platforms, and one type is a serious fault, such as: the machine temperature is too high, bus voltage is abnormal, and the like, and an inverter connected with the photovoltaic unit can stop running immediately and needs to be shut down for maintenance. The other is a general failure such as: the fan fault, etc. can not produce the trouble that influences greatly to personal and dc-to-ac converter safety, the output of photovoltaic unit only can appear to reduce to need not shut down immediately, because from discovery to maintenance still need a period of time, can practice thrift the cost.

Disclosure of Invention

In order to solve the defects in the prior art, the invention discloses a three-state sequential Monte Carlo calculation method of a power supply reliability index, in particular to a three-state sequential Monte Carlo calculation method of the power supply reliability index in the self-consistent energy system planning of a highway, which comprises the following calculation scheme:

a method for calculating a power supply reliability index of a highway self-consistent energy system is characterized by comprising the following steps:

step 1: the specific structure for establishing the expressway self-consistent energy system comprises the following steps: a supply side, a demand side, and a microgrid; the supply side consists of a wind, light and energy storage power generation system; the demand side takes the expressway demand as a main body and is connected through an intermediate medium micro-grid;

Step 2: constructing a real-time power model of a supply side and a traffic load model of a demand side of a highway self-consistent energy system: sampling and determining a wind speed scene of a wind turbine generator, an illumination radiation degree scene of a photovoltaic turbine generator and a traffic load demand scene of a highway by adopting different probability density distribution through a three-state sequential Monte Carlo method; the method comprises the steps of constructing a power generation power model of a wind turbine by using a power-wind speed conversion formula of the wind turbine, constructing a power generation power model of the photovoltaic turbine by using the power conversion formula, and analyzing historical data of traffic load, wherein the historical data of the load side can be considered to be fitted into normal distribution, and load values have little difference at the same moment in different days of each quarter all the year round, and can be described by adopting the same normal distribution. The invention adopts a typical load daily fitting method based on normal distribution to approximate load scenes of four seasons of one year, considers that the load of 24 hours of one day of the expressway approximately obeys 24 groups of different normal distribution, and obtains an expected value of the normal distribution in 24 hours per hour as a demand value of traffic load at the moment by sampling as a power model of a demand side.

Step 3: establishing a reliability parameter model of each component part of the expressway self-consistent energy system;

step 4: and calculating the power supply reliability index by using the three-state sequential Monte Carlo.

The invention discloses a highway self-consistent energy system, which is characterized in that: the system comprises a non-volatile storage medium comprising a stored program, wherein the program when run controls a device in which the non-volatile storage medium resides to perform the method described above.

The invention discloses an electronic device, which is characterized by comprising a processor and a memory; the memory has computer readable instructions stored therein, and the processor is configured to execute the computer readable instructions, where the computer readable instructions execute the method described above when executed.

Advantageous effects

When the sequential Monte Carlo of the expressway self-consistent energy system is used for calculating the power supply reliability index, uncertainty of two ends of supply and demand is fully considered when a power generation power model of a supply side and a load model of a demand side are established. Probability density distribution is respectively constructed at the supply and demand ends, li Yongmeng terCarlo samples are used for simulating a wind-light load scene, fixed historical data is not adopted, and the uncertain characteristics of a supply side and a demand side are reflected;

From the three-state sequential monte carlo algorithm itself, in the computation flow of the conventional sequential monte carlo algorithm, the elements have only two states: and the relationship between the power supply and the load is analyzed only in the two periods of normal operation time T1 and fault repair time T2, and it is considered that all kinds of elements can only stop and fault when the power supply or other switching elements are in fault, and the actual conditions of the power plant operation are not completely matched. The simulation time of the three-state sequential Monte Carlo provided by the invention comprises three states: normal operation time t1+with failure operation time t2+repair time T3 of failed device. The fault state of the power supply element is distinguished from the normal switching element by adding a period of time T2 with fault. By a random number x _j Compared with the fault probability, the fault diagnosis time T2 and the repair time T3 are determined, the photovoltaic, fan and energy storage system on the supply side are artificially set to be in the fault diagnosis time T2, the power generation capacity is reduced to 60% of the normal power generation capacity (the normal power obtained by the real-time wind speed/illumination radiation degree), and the shutdown maintenance is only carried out in the repair time T3 of the corresponding fault equipment, so that the economical idea of actual operation is ensured. And only the common switching element without generating capacity constantly shows a shutdown fault in the fault diagnosis T2 and the repair time T3.

In general, the three-state sequential Monte Carlo adopted by the invention not only considers more comprehensively and specifically from application objects, but also can embody the distinction of power supply elements and other element types from the algorithm itself, and considers the types of various faults, thereby improving the limitation that the traditional sequential Monte Carlo algorithm does not distinguish the power supply faults from other element faults such as switching elements and the like, and ensuring the economical idea in actual operation.

Drawings

FIG. 1 is a histogram of Weibull probability density distributions for wind speeds;

FIG. 2 is a schematic diagram of probability distribution of illumination radiance beta;

FIG. 3 is a graph of probability distribution of illumination irradiance beta;

FIG. 4 is a schematic diagram of a normal distribution of loads;

FIG. 5 is a simulation flow diagram of a three-state sequential Monte Carlo;

FIG. 6 is an active power output curve of a wind turbine;

FIG. 7 is a schematic diagram of a two-state model of a component;

FIG. 8 is a simplified block diagram of an energy storage device;

FIG. 9 is a state transition diagram of the energy storage device;

fig. 10 is a schematic diagram of a state transition process of a micro grid under a Well-rolling theory;

FIG. 11 is a three-state sequential Monte Carlo calculation flow chart for a power supply reliability index;

FIG. 12 is a block diagram of a highway self-consistent energy system configuration;

FIG. 13 is a schematic diagram of a highway self-consistent energy system architecture;

fig. 14 is a two-state sequential monte carlo calculation flowchart for the power supply reliability index.

Detailed Description

step 1: the specific structure for establishing the expressway self-consistent energy system comprises the following steps: a supply side, a demand side, and a microgrid; the supply side consists of a wind, light and energy storage power generation system; the demand side takes the expressway demand as a main body and is connected through an intermediate medium micro-grid.

Step 2: constructing a real-time power supply model of a self-consistent energy system supply side of the expressway and a traffic load model of a demand side: sampling and determining a wind speed scene of a wind turbine generator, an illumination radiation degree scene of a photovoltaic turbine generator and a traffic load demand scene of a highway by adopting different probability density distribution through a three-state sequential Monte Carlo method; the method comprises the steps of constructing a wind turbine generator power generation model by using a power-wind speed conversion formula of the wind turbine generator, constructing a photovoltaic generator power generation model by using the power conversion formula, adopting a scene of normal distribution construction, and constructing a traffic load model by taking an expected value of normal distribution of each hour in 24 hours as a traffic load demand value at the moment.

The Monte Carlo simulation method is a computer random simulation method based on a probability statistical theory. Monte Carlo simulation methods are classified into non-sequential and sequential Monte Carlo simulation methods. When evaluating the reliability of a power system, the sequential Monte Carlo simulation method is generally used as a unit of hours, and can simulate the state of the system over a certain time span.

Before calculating the power supply reliability index of the expressway self-consistent energy system by using a three-state sequential Carlo simulation method, the power generation side of the expressway self-consistent energy system needs to be built: the power models of the wind turbine generator, the photovoltaic turbine generator and the energy storage system also comprise a traffic-side load model and a reliability model of each element of the expressway self-consistent energy system.

1. Real-time power model of supply side

(1) Wind power model

The power curve change of the wind turbine generator is related to the change of wind speed. The most widely used today is to approximate the output power of a wind turbine by a quadratic function, as shown in fig. 5-6.

Knowing the wind speed v at a certain moment _t The output power of a single fan can be obtained by using a power conversion formula, wherein the conversion formula is as follows:

wherein: p (P) _r Rated power of a single fan; p (P) _t The power of the fan at the moment t; v _r 、v _ci 、v _co Respectively represent rated wind speed and cut of the fanWind in and wind out; A. the value of B, C depends on v _r And v _ci Is of a size of (a) and (b).

(2) Photovoltaic power model

The output power of the photovoltaic unit is related to various factors such as illumination radiance, ambient temperature, inclination angle of a photovoltaic panel, conversion efficiency of the photovoltaic unit and the like, wherein the illumination radiance has the greatest influence on the photovoltaic output power, so that when a power model is constructed in the patent, only the influence of the illumination radiance on the output power of the photovoltaic unit is considered.

When the illumination radiation degree G at a certain moment is known _t The output power corresponding to the photovoltaic unit can be obtained by using a power conversion formula, wherein the conversion formula is as follows:

wherein: p (P) _t The power of the photovoltaic unit at the moment t; p (P) _m The rated power of the photovoltaic unit is set; g _std The irradiation power of illumination given for standard environment is usually 1kW/m2; r is R _c For a specific illumination emittance, 0.15kW/m2 is usually taken.

(3) Energy storage charge/discharge model

The working principle of the energy storage system in the micro-grid system is that the energy storage system stores electric energy when wind and light supply at the power generation side is larger than load demand; when the wind-solar supply at the power generation side is smaller than the load demand, the energy storage system releases electric energy. Energy storage systems are of a wide variety including supercapacitor energy storage, battery energy storage, electrochemical energy storage, compressed air energy storage, and the like. The invention uses the accumulator as the energy storage system.

The electric energy stored in the energy storage system at the time t is B (t), and the charge/discharge power is P _B (t), the charging/discharging time sequence of the energy storage system is:

B(t+1)＝B(t)+P _B (t) (3-4)

the power limit of the energy storage device during charging/discharging is as follows:

wherein: p (P) _B The value of (t) being positive indicates that the energy storage device is charged; p (P) _B A negative value of (t) indicates that the energy storage device is discharging; p (P) _disch-max Maximum discharge power of the energy storage device; p (P) _ch-max Maximum charging power for the energy storage device; b (B) _max Is the maximum capacity of the energy storage device; b (B) _min Is the minimum capacity of the energy storage device.

2. Traffic energy load model on demand side

Aiming at a load curve of a self-consistent energy system on a highway on a demand side, after load data of a traffic demand side is comprehensively analyzed, the load values of different days at the same moment are considered to be not greatly different, so that a typical load day fitting method based on normal distribution is adopted. The load side of the highway is considered to be approximately compliant with a normal distribution, and the desire for a normal distribution per hour over 24 hours is considered to be the traffic load demand at that moment.

The normal distribution model is a probability distribution model which is frequently used in the engineering field. The normal distribution is also called gaussian distribution, and its functional image is called normal curve. If the probability density function f (Z) of the random variable Z is

Wherein: sigma > 0, and mu, sigma are constants, called random variables, Z obeys the parameters mu, sigma ² Is denoted as Z to N (μ, σ) ² ). To fit random variables to normal distributions, only μ and σ need to be determined ² I.e., where μ is the expectation of the distribution, σ ² Is the variance of the distribution.

The typical load day fitting method divides 1 day into 24 time periods according to 1h as 1 time period, then utilizes a large amount of power grid load data to fit the load data at the same moment in different days into normal distribution, and finally takes the expected normal distribution at each moment as the load value at the moment in the typical day. The final fitted typical day is obtained by calculating the load values for 24 time periods.

After setting the normal distribution function of the ith period in 24 hours by the formula (3-7), the expected load value of the ith period is calculated by the following formula

E(x _i )＝∫x _i f(x _i )dx _i (3-8)

Wherein: e (x) _i ) Indicating the desire for the ith period; x is x _i An argument representing an i-th period normal distribution function, f (x _i ) Representing the normal distribution function of the i-th period.

The expected value is used as the load value at the moment, and after all the moment expectations are calculated, the expectations are used to fit the typical load day, namely E (x) ₁ )，E(x ₂ )，…，E(x ₂₄ ) As the load value for the typical load day fitted.

Step 3: and establishing element reliability parameter models of all components of the highway self-consistent energy system, wherein the element reliability parameter models comprise a reliability model of an energy storage device, a reliability model of a wind turbine generator system and a reliability model of a photovoltaic turbine generator system.

The reliability research of the invention considers that elements in the element power system have three states of normal operation, fault and maintenance. The operational state of the elements in the system is simulated by means of a conventional markov model, as shown in fig. 7.

Mean normal operation time of the element is T _F Average repair time T _R Assuming that the state duration obeys an exponential distribution, T _F And T _R The expression is:

wherein: x is x ₁ And x ₂ Is a random number uniformly distributed in (0, 1) and can be generated by using a rand () function in Matlab; λ and μ represent the equivalent failure rate and the equivalent repair rate of the element, respectively.

Reliability model of energy storage device

The components in the microgrid may be divided into power components and non-power components, where the power components have: wind turbine generator system, photovoltaic unit and energy storage system, non-power supply element mainly has: circuit breakers, transformers and fuses. The reliability data for the failure rate and repair rate of the non-power supply element is obtained by referring to the relevant literature. The invention adopts a state space method to calculate the equivalent fault rate and the repair rate of the wind turbine generator, the photovoltaic turbine generator and the energy storage system.

The state space method is to represent the system by its state and possible transition between its states, and to find the equivalent failure rate and repair rate of the system based on this. The method comprises the following specific steps: listing all possible states of the system; forming a state transition diagram and a state transition matrix; determining the probability of the stable state of the system according to the failure rate and the repair rate of each component element in the system; and calculating the equivalent failure rate and the equivalent repair rate of the whole system.

The energy storage system stores energy for the storage battery. The energy storage device reliability model only considers the storage battery double-stage structure formed by the series connection of the storage battery pack, the DC/DC converter, the DC/AC inverter and the grid-connected filter, and the structure is shown in figure 8. The discharging process of the storage battery in the structure is that direct current is converted by a DC/DC converter and a DC/AC inverter and is integrated into a large power grid through a three-phase filter.

Let the failure rate of each part of the energy storage system be lambda _C1 、λ _C2 、λ _C3 、λ _C4 Repair rates were μ, respectively _c1 、μ _c2 、μ _c3 Sum mu _c4 Equivalent failure rate lambda of energy storage system _c The method comprises the following steps:

λ _c ＝λ _c1 +λ _c2 +λ _c3 +λ _c4 (3-10)

the step of solving the equivalent repair rate of the energy storage device is as follows:

(1) Since the four parts constituting the energy storage device belong to the series system, any part fails, the whole system stops operating and is in a failure state, and only the first order failure is considered, thereby forming a state transition diagram of the energy storage device, as shown in fig. 9.

In fig. 9, numeral 0 indicates that the apparatus is in a normal operation state, and

numerals

1, 2, 3, and 4 respectively indicate apparatus failure states caused by the battery pack, the DC/DC converter, the DC/AC inverter, and the filter.

(2) A state transition matrix a is formed. A is:

(3) Solving stationary state probability p= [ p ] of energy storage device ₀ ,p ₁ ,p ₂ ,p ₃ ,p ₄ ]Wherein p is ₀ For the probability of normal state of the energy storage device, p ₁ ,p ₂ ,p ₃ ,p ₄ The failure probabilities of the various parts of the energy storage device are respectively. The equation set for solving p is:

(4) According to p ₀ Relationship with equivalent failure rate lambda and equivalent repair rate mu, solving equivalent repair rate mu _c . Wherein p is ₀ And lambda is _c 、μ _c The relationship of (2) is as follows:

equivalent repair rate mu of energy storage device _c The method comprises the following steps:

in summary, the failure rate lambda of each of the four parts of the energy storage device is obtained by searching the literature and related data _C1 、λ _C2 、λ _C3 、λ _C4 And repair rate mu _c1 、μ _c2 、μ _c3 Sum mu _c4 The energy storage device reliability model established in the part is utilized, and the equivalent failure rate lambda of the energy storage device can be obtained by utilizing formulas (3-10) and (3-14) _c (times/year) and equivalent repair rate mu _c (times/hour).

Reliability model of wind turbine generator

When the reliability model of the wind turbine is built, only four parts of the wind driven generator, the AC/DC rectifier, the DC/AC inverter and the filter are considered to build the reliability model of the wind turbine. Taking a two-stage structure of a permanent magnet direct-drive wind generating set as an example, a simplified wind generator model is adopted. The wind driven generator is connected to a bus of the micro-grid through an AC/DC rectifier, a DC/AC inverter and a filter. And obtaining the failure rate and the repair rate of each part by searching documents and related data, and popularizing the reliability model of the energy storage device to the wind turbine generator.

The failure rate of four components of the wind turbine generator is lambda _F1 、λ _F2 、λ _F3 、λ _F4 Repair rates were μ, respectively _F1 、μ _F2 、μ _F3 Sum mu _F4 Equivalent failure rate lambda of wind turbine generator _F The method comprises the following steps:

λ _F ＝λ _F1 +λ _F2 +λ _F3 +λ _F4 (3-15)

equivalent repair rate mu of wind turbine generator _F The method comprises the following steps:

reliability model of photovoltaic unit

The topology of the photovoltaic unit is composed of two types: two-stage structures and single-stage structures. The double-stage structure is that a photovoltaic array raises or lowers direct current voltage through a DC/DC converter to reach voltage required by inversion, and meanwhile Maximum Power Point Tracking (MPPT) is realized, then the direct current is converted into alternating current through the DC/AC inverter, and the alternating current is filtered through a filter and is combined on a bus of a micro-grid.

In the invention, the photovoltaic unit adopts a two-stage structure, and a reliability model of the photovoltaic unit is built only by considering the four parts of the photovoltaic array, the DC/DC converter, the DC/AC inverter and the filter. And obtaining the fault rate and the repair rate of each part through documents and related data, and popularizing the reliability model of the energy storage device to the photovoltaic unit.

The failure rate of the four components of the photovoltaic unit is lambda _G1 、λ _G2 、λ _G3 、λ _G4 Repair rates were μ, respectively _G1 、μ _G2 、μ _G3 Sum mu _G4 Equivalent failure rate lambda of photovoltaic unit _G The method comprises the following steps:

λ _G ＝λ _G1 +λ _G2 +λ _G3 +λ _G4 (3-17)

equivalent repair rate mu of photovoltaic unit _G The method comprises the following steps:

The biggest characteristic of the expressway self-consistent energy system is uncertainty of both supply and demand ends. Because the natural resources of wind and light at the supply side are greatly influenced by factors such as surrounding environment, weather and the like when generating electricity, the output has intermittence and fluctuation. Meanwhile, the traffic load on the demand side is also affected by factors such as weather, environment and the like, and the uncertainty is also caused. Therefore, when the power supply reliability index of the highway self-consistent energy system is calculated, uncertainty of wind, light and load needs to be considered, and probability density distribution can be constructed respectively.

1. The application of the monte carlo method was analyzed.

(1) By using the Monte Carlo method, firstly, a statistical experiment probability model which is matched with the actual problem is constructed by combining the actual physical property of the problem, the model is properly adjusted, and proper probability model parameters are selected.

(2) When sampling from a matrix with known distribution, it is necessary to take into account the method of sampling from a matrix with known distribution according to the distribution of each random variable in the model, and generate sufficient random numbers in computer simulation.

(3) For different models, no exact formula can directly determine the optimal number of simulated scenes. The field Jing Su subtraction is typically used, i.e. through a large number of simulation experiments, the minimum number of scenes is chosen when the simulation results are substantially stable and unchanged.

2. Simulation of wind and light load scene by Monte Carlo sampling

After summarizing the respective properties of wind, light and traffic load, the invention adopts Weibull probability density distribution to construct wind speed, adopts beta distribution to construct illumination radiation intensity, and adopts normal distribution to construct traffic demand end load model.

The wind speed, the illumination radiance and the system load can meet the requirements of weibull, beta and normal distribution in h, the wind speed of 24h a day, the 24 groups of probability density distribution of the illumination radiance and the system load are respectively set, scene sampling is carried out by utilizing Monte Carlo, and a scene of wind and light load under the condition of considering uncertain factors is obtained.

(1) The wind speed accords with Weibull probability distribution every hour, the illumination radiance accords with beta distribution at h, and the load accords with normal distribution at the same time of different days.

(2) And simulating and generating wind speeds, illumination radiance and loads of multiple scenes by using matlab functions. (3) Scene cuts may be made under conditions that satisfy the distribution density function.

Probability density distribution of wind speed

Wind speed is a random phenomenon, the distribution form of which can be approximately analyzed, and the probability model of wind speed adopts Weibull distribution in the invention. The Weibull probability density function is as follows:

wherein: c. k is a scale parameter (m/s) and a shape parameter of the model respectively; v is wind speed (m/s). If the wind speed average v and the wind speed variance sigma are known, the calculation formulas of the parameters c and k can be obtained as follows:

wherein:

σ is the mean value (m/s) and standard deviation (m/s) of the wind speed history data, respectively.

Integrating the formula (2-1) to obtain a probability distribution function F (v) of the wind speed as follows:

let u=f (v), r=1-U, and find the wind speed v by inverting the equation (2-4) _t The method comprises the following steps:

to sum up, after obtaining parameters c and k of Weibull distribution model, for any moment, utilizing rand in Matlab) The function generates a random number R belonging to (0, 1) and brings the random number R into the formula (2-5) to obtain the wind speed simulation value v _t 。

When the number of the sampling generation scenes is set to be 1000, the shape parameter k of the Weibull model of the wind speed is 1.637; taking 5.218 as a scale parameter c; the Weibull probability distribution of wind speed at this time is shown in FIG. 1.

Probability distribution of illumination radiation intensity

The influence factors of the illumination radiation intensity are clouds, shadows and the like, and the probability distribution of the illumination radiation intensity within 1 hour or a plurality of hours can be approximated as Beta probability distribution. The Beta distribution is a set of continuous probability distributions defined over the interval [0,1], with two shape parameters alpha and Beta, the probability density function being

Wherein r is solar irradiance at a certain time in the period, W/m ² The method comprises the steps of carrying out a first treatment on the surface of the rmax is the maximum irradiance of the sun, W/m, for this period ² The method comprises the steps of carrying out a first treatment on the surface of the Γ () is a gamma function; alpha, beta are shape parameters of the Beta distribution, the variation of which will lead to the variation of the shape of the Beta distribution probability density curve, and alpha, beta can be calculated from the mathematical expectation mu and variance delta of the solar radiation intensity over a period of time:

when the number of the sampling generation scenes is set to be 1000, the shape parameters of the beta model of the illumination radiance; the shape parameters α= 0.6869, β= 2.1320, and the irradiance beta probability distribution diagram at this time is shown in fig. 3.

Probability distribution of load model

Aiming at a load curve of a self-consistent energy system on a highway on a demand side, after load data of a traffic demand side is comprehensively analyzed, the load values of different days at the same moment are considered to be not greatly different, so that the patent adopts a typical load day fitting method based on normal distribution. The load side of the highway is considered to be approximately compliant with a normal distribution, and the desire for a normal distribution per hour over 24 hours is considered to be the traffic load demand at that moment.

The normal distribution model is a probability distribution model which is frequently used in the engineering field. The normal distribution is also called gaussian distribution, and its functional image is called normal curve. If the probability density curve of the random variable Z is

Wherein sigma is more than 0, mu and sigma are constants, the random variable is called, and the obeying parameter of z is mu and sigma ² Is denoted as Z to N (μ, σ) ² ). To fit random variables to normal distributions, only μ and σ need to be determined ² I.e., where μ is the expectation of the distribution, σ ² Is the variance of the distribution.

When the number of sampling generation scenes is set to be 1000, in normal distribution of the load model, the average value mu=5.91e3w; variance σ2=0.1 μ. The normal distribution of the load at this time is shown in fig. 4.

The traditional sequential Monte Carlo simulation fault is based on a two-state reliability model, monte Carlo sampling is carried out on the fault state and the fault time of the element in the time of the simulation, the relation between the power supply and the load in the normal operation time and the fault state influence range of the element is analyzed, the relevant parameters of the load are accumulated, and finally the reliability index of the load is utilized to evaluate the reliability of the system. The simulation steps are as follows:

(1) An indicator of device reliability (failure rate, repair rate) is determined.

(2) After the simulation is started, calculating the normal working time T of each device _i And determining the failed device by setting certain rules.

Typically lambda is set _i For the failure rate of element i (i=1, 2, …, n), there are n elements in the system. Generating n random numbers x using function rand () in Matlab _i Calculating the normal running time T of each element _i ＝-(ln x _i )/λ _i The method comprises the steps of carrying out a first treatment on the surface of the Let T ₁ ＝min(T _i )，T ₁ For minimum normal operation time, T ₁ The corresponding element i fails.

(3) And determining the repair time of the fault equipment.

Regenerating 1 random number x _j Calculating fault repair time Tmu of faulty element j _j Wherein Tμ _j ＝–(ln x _j )/μ _j ，μ _j Is the repair rate of the failed element j.

(4) And determining the load which is stopped due to equipment faults, and accumulating the parameters such as normal working time, stopping times, stopping time and the like of the load.

(5) Simulated time, plus this Monte Carlo simulation time: and (5) judging whether the simulation threshold is reached or not (minimum normal running time and fault repair time), if so, carrying out the next step, otherwise, returning to the step (2), and then carrying out the next sequential Monte Carlo simulation.

(6) Statistical load reliability index and system reliability index.

Because the traditional sequential Monte Carlo simulation considers that each element loses the capacity of the power supply, the switching element and the line immediately after the element fails, the power supply, the switching element and the line are in a completely failed open state, and the operation time of one sequential Monte Carlo only comprises the normal operation time and the fault repair time. In fact, a period of time after any fault occurs is the fault diagnosis analysis time. And when some small faults occur to the generating set such as wind and light, the generating set does not appear as a shutdown fault, and instantly loses the power generation capacity, but appears as weakening of the power generation capacity.

The three-state sequential Monte Carlo simulation time adopted by the invention comprises three parts: normal operation time t1+with failure operation time t2+repair time T3 of failed device. Thus, the power supply and the common switching element are distinguished.

The traditional sequential Monte Carlo ignores the analysis of the types of the elements, and does not distinguish the power failure from other element failures such as a switching element, while the three-state sequential Monte Carlo can embody the distinction of the power element from other element types.

The simulation procedure for the three-state sequential Monte Carlo is as follows:

Typically lambda is set _i I (i=1, 2, …, n). Generating n random numbers x using function rand () in Matlab _i N is the number of all elements. Calculating the normal operation time T of each element _i ＝-(ln x _i )/λ _i The method comprises the steps of carrying out a first treatment on the surface of the Let T ₁ ＝min(T _i )，T ₁ For minimum normal operation time, T ₁ The corresponding element i fails.

(3) Repair time T3 with failed run time t2+ failed device.

A random integer x of 1 to n is generated by using a function rand () in Matlab,

If x is not less than 1 and not more than a, determining that the element in the power element fails, wherein a is the sum of numbers of the wind driven generator, the photovoltaic unit and the energy storage system. If wind power, photovoltaic and energy storage system have faults, 1 random number x of 0-1 is correspondingly generated _j Calculating the sum T of the repair time T3 of the fault-carrying running time T2+ fault equipment of the fault element j _μj Wherein T is _μj ＝-(ln x _j )/μ _j ，μ _j Is the repair rate of the failed power supply j.

Wherein the wind power, the photovoltaic power and the energy storage system are represented in the fault diagnosis time T2 that the power generation capacity is reduced to 60% of the normal power generation capacity (normal power obtained by real-time wind speed/illumination radiation), and T2 occupies 2/3 of T _μj Time, but only at repair time T of the corresponding faulty device ₃ ＝T _μj -T ₂ And (5) stopping and overhauling the machine.

If a+1 is not less than x is not less than n, other switching elements, circuit breakers, transformers and the like are failed: during the repair time T3 of the device with fault run time t2+, it is shown as a fault shutdown.

(5) Simulated time, plus this Monte Carlo simulation time: and (3) judging whether the normal working time T1+ the fault running time T2+ the repair time T3 of the fault equipment reaches a simulation threshold value, if so, performing the next step, otherwise, returning to the step (2) and then performing the next sequential Monte Carlo simulation.

(6) Statistical load reliability index and system reliability index.

The conventional sequential monte carlo ignores the analysis of the types of the elements, does not distinguish the power failure from other element failures such as switching elements, and does not consider the failure type of the element. The three-state sequential Monte Carlo not only can embody the distinction between the power supply element and other element types, but also can consider various fault types.

The Well-rolling theory is applied to calculation of the power supply reliability index of the highway self-consistent energy system.

(1) Application of Well-rolling theory in micro-grid

The Well-rolling theory divides the power generation system into three states: health status, boundary status, and risk status; each state is represented by a probability.

At a certain moment, the health state indicates that the power generation side supply amount of the micro-grid is larger than the demand amount of the load at the current moment, and has a certain standby capacity, and the standby capacity of the health state is set to be 0.1 times of the maximum load with reference to the traditional power system. The boundary state indicates that the supply amount on the power generation side at the present time is exactly equal to the demand amount of the load. The risk state indicates that the power generation side supply amount cannot satisfy the load demand amount at the present time, and that a part of the load is powered off.

Since the wind power and photovoltaic power generation on the power generation side are greatly affected by a plurality of factors such as external environment, when the micro-grid independently operates, the operating state thereof is converted between a healthy state, a boundary state and a risk state, as shown in fig. 10. And evaluating the reliability degree of the micro-grid power generation system through probability values of the three states.

(2) Calculation formula of power supply reliability index

The expressions that represent that the micro-grid is in a healthy state, a boundary state and a risk state are as follows:

/>

wherein: s is S _H Indicating that the microgrid is in a healthy state; p (P) _Lmax Representing the maximum load of the micro-grid; p (P) _Lt Representing the load of the micro-grid at the time t; p (P) _t The maximum power which can be output by the micro-grid at the moment t is represented; s is S _M Indicating that the micro-grid is in a boundary state; s is S _R Indicating that the microgrid is in a risk state.

The reliability evaluation takes 1 hour as a step length, the output power and the load of the micro-grid are considered to be constant within 1 hour, and the indexes for representing the reliability of the micro-grid power generation system by using the Well-rolling theory are as follows: health state probability P (H), boundary state probability P (M), risk state probability P (R). The calculation expression of each index is as follows:

wherein: n represents the simulated years of the reliability evaluation process of the power generation system; n (H), n (M), n (R) represent cumulative hours of the microgrid being in a health state, a boundary state, and a risk state during reliability evaluation, respectively.

In combination with the definition of the power supply reliability index in the dl_t5542-2018 industry standard: the power supply Reliability (RS) represents the ratio of the total number of hours of power supply time available to the user to the number of hours of the statistical period, denoted as RS, and can be calculated as follows:

wherein: t (T) _user Represents average power failure time, T _st Representing the time during which statistics were made.

In conclusion, based on the Well-rolling theory and the DL_T5542-2018 industry standard, the calculation formula of the power supply reliability index of the highway self-consistent energy system is as follows

(3) Flow for calculating power supply reliability index by using three-state sequential Monte Carlo method

The wind, light and storage form a supply side, the expressway is used as a demand side, and the micro-grid is used as an intermediate framework. The strategy of the energy storage system in the expressway self-consistent energy system is as follows: and when the output force of the supply side is larger than the load demand at the moment t, the energy storage device is charged. When the supply side output is less than the load demand, the energy storage device discharges. The system fault condition is simulated by utilizing the three-state sequential Monte Carlo, and the single-order element fault is considered, so that the specific flow of calculating the power supply reliability index is as follows: (1) and constructing and establishing a power model of a supply side and a demand side of the highway self-consistent energy system. And constructing a wind speed Weibull distribution model parameter and an illumination radiance Beta model, sampling N=1000 wind-light scenes of 8760 hours by using Monte Carlo, and obtaining the output power of the unit under the wind-light scenes of 8760 hours by using a power conversion formula. And the load side adopts normal distribution to construct a probability model, and N=1000 load scenes of 8760h are sampled by using Monte Carlo. (2) And establishing a reliability parameter model of each component of the highway self-consistent energy system. The three-state sequential Monte Carlo method considers that the operational state of all elements consists of three sub-states: normal run time T1, with fault run time T2, and fault repair time T3. The specific time of three states of each element is acquired, so that the equivalent failure rate lambda and the equivalent repair rate mu of all the component elements needing to consider the failure condition in the system can be clarified, and specific parameters can be obtained by referring to related documents for the failure rate lambda and the repair rate mu of the circuit breaker of the key equipment of the micro-grid; for the wind power, photovoltaic and energy storage integrated system, the internal mechanism is complex and various, and a simplified two-stage structure is adopted. Firstly, obtaining the failure rate lambda and the repair rate mu of the internal components of the double-stage structure by referring to related documents, and then calculating the equivalent failure rate lambda and the equivalent repair rate mu of the wind power, photovoltaic and energy storage integrated system by using a state space method.

(3) After the power model and the reliability model are established, simulation can be performed. Setting the total years n=1000, the simulation start years mt=1, ht=0 of the simulation means starting the simulation from 0 hours up to 8760 hours. All elements are in a normal state at the beginning; the micro-grid considers the fault condition of the breaker element, configures the capacity and the number of the wind-solar energy storage systems and formulates a strategy of the energy storage charging/discharging process. In the simulation process, the supply and demand relation of the system is compared, and reliability parameters provided by a well-rolling theory are utilized: the cumulative hours n (H), n (M), and n (R) of the health state, the boundary state, and the risk state, and the load cumulative reduction L are accumulated per hour. A group of 8760h wind-light load data is read in to perform three-state sequential Monte Carlo simulation, and all elements are in a normal state at the beginning moment: n (H) =0, n (M) =0, and n (R) =0, and the load accumulation reduction amount l=0.

(4) First, the element that has failed is determined by the minimum normal operation time T1. Setting the number of all elements in the system as n, generating n random numbers x by utilizing Matlab function rand () _i Calculating the normal working time T of each power supply _i ＝-(ln x _i )/λ _i ，λ _i For the failure rate, x, of each element i (i=1, 2, …, n) in the system _i Is a random number which is uniformly distributed according to 0-1. Let T ₁ ＝min(T _i ) T is then ₁ The corresponding power source i fails, and the element i with the shortest normal operation time is considered to fail first.

(5) The repair time T3 of the failed component with the failed run time t2+ failed device is determined.

And generating a random integer x of 1-n by using a function rand () in Matlab. If x is not less than 1 and not more than a, determining that the power supply element, namely the wind driven generator, fails, x is an integer random number which accords with 1-n uniform distribution, a is the sum of the numbers of the elements of the wind driven generator, the photovoltaic unit and the energy storage system, and n is the total number of all the elements in the system. When the power supply element failsIn this case, 1 random number x is generated _j Calculating the sum T of the repair time T3 of the fault-carrying running time T2+ fault equipment of the power supply element j _μj Wherein T is _μj ＝-(ln x _j )/μ _j ，μ _j For repairing rate, x of faulty power supply element j _j Is a random number which is uniformly distributed according to 0-1. The artificial setting of the fault power supply element is represented by the reduction of the power generation capacity to 60% of the normal power generation capacity (the normal power obtained by the real-time wind speed/illumination radiance through a power conversion formula) within the fault diagnosis time T2, and T2 occupies 2/3 of T _μj Time, but only at repair time t3=t of the corresponding faulty device _μj -performing a shutdown service within T2.

If a+1 is less than or equal to x is less than or equal to n, other switch elements, circuit breakers, transformers and the like are failed, and the elements do not have the power generation capacity but only take the function of transmitting electric energy, so that after the elements are failed, the elements are all in failure stop in the repair time T3 of equipment with the failure running time T2 < + >.

(6) And carrying out power balance analysis on the system in the time period from HT to HT+T1+T2+T3 (starting time h to starting time h+normal working time T1+repairing time T3 of the equipment with fault running time T2). Let t=0:

1) t '=ht+t, and according to the power model and the load model of wind power, photovoltaic and micro power supply, calculating the wind power and the light power at the time t' as P respectively _W (t')、P _S (t') load power P _L (t'). The maximum power that the microgrid can output at this time is P (t')=p _W (t')+P _S (t')+P _B (t')(P _B And (t ') is the maximum discharge power of the energy storage device at the time t').

2) If P _W (t')+P _S (t')≥P _L (t'), charging the energy storage device, and executing step 4); otherwise go to step 3).

3) If P _W (t')+P _S (t')+P _B (t')≥P _L (t'), then executing step 4); otherwise l=l+p _L (t ') -P (t'), n (R) =n (R) +1, go to step 5).

4) If P (t')ismore than or equal to 1.1P _Lmax N (H) =n (H) +1, go to step 5); otherwise n (M) =n (M) +1, go to step 5).

5) If t<T _λj +T _μj -1, then t=t+1, returning to step 1); otherwise, executing the step (6).

(7) Advancing the simulation time to ht=ht+t _λj +T _μj If HT<8760, returning to step (3); otherwise mt=mt+1 is performed.

The above process is repeated, no more than 8760 hours are required for each hour of forward progress, and once 8760 hours are exceeded, the above cycle is terminated, and the process proceeds to step (7) to determine whether to restart fault simulation for a new year 8760 hours.

(8) And when MT is less than or equal to N, reading 7680h data of a second group of wind-light load, returning to the step (3), and performing fault simulation of the next year. Otherwise, the loop is exited, and the accumulated hours of the health state, the boundary state and the risk state of the accumulated data are n (H) =0, n (M) =0 and n (R) =0; load cumulative reduction amount l=0; calculating power supply reliability index of expressway self-consistent energy system

To sum up, the power supply reliability index calculation flow of the expressway self-consistent energy system is shown in the following fig. 11.

Examples

Firstly, before the calculation of the power supply reliability index of the highway self-consistent energy system, the data needed to be used are as follows:

1. the uncertain wind-light load scene is constructed, and the required data are as follows:

(1) The average value and data of wind speeds of 24h per hour for a typical day of four quarters are used to construct the dimensional parameters c and shape parameters k of the weibull distribution.

(2) The average and variance of the illumination emittance per hour in the typical solar illumination emittance 24h of four quarters are used to construct Beta distribution shape parameters alpha, beta.

(3) The expected and variance of the load per hour for a typical daily load model 24h for four quarters is used to construct a normal distribution of load.

2. Because the system considers two fans, unit parameters of the fans need to be acquired respectively: cut in wind speed, cut out wind speed and rated wind speed; the capacity and the number parameters of wind power, photovoltaic and energy storage units; maximum/small state of charge, maximum amplification/charge power of the energy storage unit.

3. Reliability parameters of wind power, photovoltaic and energy storage systems, and other important equipment and switching elements of micro-grids (failure rate lambda of all elements in the system _i Repair rate mu _j )。

In the simulation, an uncertain wind-light load scene is constructed, a wind speed probability density function of four quarters of different typical days 24h per hour is constructed by Weibull probability density, an illumination radiance probability density function of four quarters of different typical solar radiance 24h per hour is constructed by Beta model, and a probability density function of four quarters of loads of different typical daily load models 24h per hour is constructed by normal distribution.

The system considers two fans, and the unit parameters of the fans are as follows: the cut-in wind speed, cut-out wind speed and rated wind speed, power supply capacity and reliability parameter data configuration are shown in fig. 12 and table 3-1. The wind power, photovoltaic and energy storage unit adopts a simplified unit structure, and the equivalent failure rate and the band repair rate of the whole system are obtained by consulting reference documents.

TABLE 3-1 Power (Capacity) configuration and reliability parameters for Power supplies

/>

The micro power grid considers two different fault repair times of the wind power unit, the photovoltaic unit, the energy storage unit and the circuit breaker, and the micro power source DG and the public load can be connected or disconnected with the power grid through the respective circuit breaker according to the running requirement of the micro power grid system. The constructed micro-grid structure is shown in fig. 13:

with the configuration of fig. 14 and 13 and table 3-1, the key characteristic index evaluation calculation carried into fig. 11 for the highway self-consistent energy system is performed as follows.

The results of the index run obtained by programming simulation using matlab are shown in table 3-2.

TABLE 3-2 Power supply reliability calculation results

6. Comparison with the two-state sequential Monte Carlo example

The input variables are brought into two-state sequential Monte Carlo simulation, and the specific flow of calculating the power supply reliability index is as follows:

(1) According to wind-light load historical data, constructing wind speed Weibull distribution model parameters; a light irradiation degree Beta model; and a normal distribution probability model of the load. N=1000 wind-light load scenes of 8760h are sampled by using Monte Carlo.

(2) And initializing parameters. Setting the total years n=1000, the simulation start years mt=1, ht=0 of the simulation means that the simulation starts from 0 hour. The cumulative hours of the health state, boundary state, risk state are n (H) =0, n (M) =0, and n (R) =0, respectively; the cumulative number of times of the health state, the boundary state, and the risk state is M (H) =0, M (M) =0, and M (R) =0, respectively; load cumulative reduction amount l=0; setting the number of the elements as n, and setting all the elements in a normal state at the starting moment; the power limit considered during the energy storage charge/discharge process. A set of 8760h wind and solar load data is read in.

(3) The failed component is determined. Let lambda set _i For the failure rate of i (i=1, 2, …, n), n random numbers x are generated using function rand () in Mat/ab _i Calculating the normal operation time T of each power supply _i ＝-(ln x _i )/λ _i The method comprises the steps of carrying out a first treatment on the surface of the Let T ₁ ＝min(T _i ) T is then ₁ The corresponding power supply i fails. (4) Determining a repair time T2, t2=t of a faulty device of a faulty element _μj ＝-(ln x _j )/μ _j ，μ _j For the repair rate of the fault power supply j, T _μj Repair time for component failure equipment, x _j Is a random number which is uniformly distributed according to 0-1.

(5) And carrying out power balance analysis on the micro-grid in the time period from HT to HT+T1+T2 (starting time h to starting time h+normal working time T1+repair time T2 of the fault equipment). Let t=0:

1) t '=ht+t, and according to the power model and the load model of wind power, photovoltaic and micro power supply, calculating the wind power and the light power at the time t' as P respectively _W (t′)、P _S (t') load power P _L (t'). The maximum power that the microgrid can output at this time is P (t')=p _W (t′)+P _S (t′)+P _B (t′)(P _B And (t ') is the maximum discharge power of the energy storage device at the time t').

2) If P _w (t′)+P _S (t′)≥P _L (t'), charging the energy storage device, and executing step 4); otherwise go to step 3).

3) If P _W (t′)+P _S (t′)+P _B (t′)≥P _L (t'), then executing step 4); otherwise l=l+p _L (t ') -P (t'), n (R) =n (R) +1, go to step 5).

5) If T is less than T _λj +T _μj -1, then t=t+1, returning to step 1); otherwise, executing the step (6).

(6) Advancing the simulation time to ht=ht+t ₁ +T ₂ Returning to step (3) if HT < 8760; otherwise mt=mt+1 is performed.

(7) And when MT is less than or equal to N, reading 7680h data of a second group of wind-light load, returning to the step (3), and performing fault simulation of the next year. Otherwise, the loop is exited, and the accumulated hours of the health state, the boundary state and the risk state of the accumulated data are n (H) =0, n (M) =0 and n (R) =0; load cumulative reduction amount l=0; and calculating the power supply reliability index of the highway self-consistent energy system.

To sum up, the power supply reliability index calculation flow of the expressway self-consistent energy system is shown in the following fig. 12.

FIG. 14 is a two-state sequential Monte Carlo calculation flowchart for a Power reliability index

With the configuration of fig. 14 and 13 and table 3-1, the key characteristic index evaluation calculation carried into fig. 14 for the highway self-consistent energy system is performed as follows.

The results of the index run obtained by programming simulation using matlab are shown in tables 3-3.

TABLE 3-3 Power supply reliability calculation results

Index name	Index abbreviations	Unit (B)	Calculation result
				Reliability of power supply	RS	％	90.0462％
Desired shortage of power	EENS	kWh/year	2872.287 kWh/year

As can be seen from comparison of the index results calculated by the two-state sequential Monte Carlo algorithm and the three-state sequential Monte Carlo algorithm of Table 3-2, since the three-state algorithm considers the power generation side with fault operation time, the power generation capacity of the generator set is deteriorated in the power generation side with fault operation time, and the maintenance cannot be stopped immediately, the calculated expected shortage power supply quantity, namely the load reduction quantity, is smaller than the load reduction quantity calculated by the two-state sequential Monte Carlo algorithm, and the power supply reliability of the system is higher. Therefore, the three-state sequential Monte Carlo algorithm is in line with the processing situation of coping with faults in an actual electric field.

Aiming at the calculation of the power supply reliability index of the expressway self-consistent energy system in the planning stage, the invention innovates a three-state sequential Monte Carlo algorithm considering multiple fault conditions of elements. When the sequential Monte Carlo of the expressway self-consistent energy system is used for calculating the power supply reliability index, uncertainty of both supply and demand ends is fully considered when a power generation power model of a supply side and a load model of a demand side are established. And constructing probability density distribution at each of the supply and demand ends, and sampling Li Yongmeng terCarlo to simulate a wind-light load scene.

From the three-state sequential monte carlo algorithm itself, in the computation flow of the conventional sequential monte carlo algorithm, the elements have only two states: and analyzing the relation between the power supply and the load only in the two periods of normal operation time T1 and fault repair time T2, wherein the power supply and the load are considered to be shut down faults only when the power supply or other switching elements are in fault. The three-state sequential Monte Carlo simulation time includes three states: normal operation time t1+with failure operation time t2+repair time T3 of failed device. Thereby distinguishing the fault state of the power supply element from the normal switching element. By a random number x _j Comparing with the fault probability, determining fault diagnosis time T2 and repair time T3, wherein the photovoltaic, fan and energy storage system at the supply side are in the fault diagnosis time T2The power generation capacity is reduced to 60% of the normal power generation capacity (normal power obtained by real-time wind speed/illumination radiation), and the shutdown maintenance is only carried out within the repair time T3 of the corresponding fault equipment. And the common switching element constantly shows a shutdown fault in the fault diagnosis T2 and the repair time T3. The three-state sequential Monte Carlo not only is more comprehensive and specific from the aspect of application objects, but also can embody the distinction between power supply elements and other element types from the algorithm itself, and the limitation that the traditional sequential Monte Carlo algorithm does not distinguish power supply faults from other element faults such as switching elements is improved by considering various fault types.

The foregoing has shown and described the basic principles, principal features and advantages of the invention. It will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, and that the above embodiments and descriptions are merely illustrative of the principles of the present invention, and various changes and modifications may be made therein without departing from the spirit and scope of the invention, which is defined by the appended claims. The scope of the invention is defined by the appended claims and equivalents thereof.

Claims

1. A method for calculating a power supply reliability index of a highway self-consistent energy system is characterized by comprising the following steps:

step 2: constructing a power model of a supply side and a traffic load model of a demand side of a highway self-consistent energy system: respectively adopting different probability density distributions, sampling and determining a wind speed scene of the wind turbine generator, an illumination radiation degree scene of the photovoltaic turbine generator and a traffic load demand scene of a highway by a three-state sequential Monte Carlo method, constructing a wind turbine generator power model by using a power-wind speed conversion formula of the wind turbine generator, and constructing a photovoltaic turbine generator power model by using a power conversion formula; and (3) adopting a typical load day fitting method based on normal distribution to approximate a load scene of four quarters of a year, and sampling to obtain an expected value of normal distribution per hour in 24 hours as a traffic load demand value at the moment to serve as a traffic load model at the demand side.

Step 3: establishing a reliability parameter model of elements of each component part of the expressway self-consistent energy system;

2. The method for calculating the power supply reliability index of the highway self-consistent energy system according to claim 1, wherein the method comprises the following steps: the power model of the real-time power supply of the supply side includes: wind power model:

the output power of a single fan can be obtained by using a power conversion formula, wherein the conversion formula is as follows:

wherein: p (P) _r Rated power of a single fan; the wind speed at a certain moment is v _t ；P _t The power of the fan at the moment t; v _r 、v _ci 、v _co Respectively representing the rated wind speed, the cut-in wind speed and the cut-out wind speed of the fan; A. the value of B, C depends on v _r And v _ci Is of a size of (2);

photovoltaic power model:

illumination radiation degree G at a certain moment _t The output power corresponding to the photovoltaic unit can be obtained by using a power conversion formula, wherein the conversion formula is as follows:

wherein: p (P) _t The power of the photovoltaic unit at the moment t; p (P) _m The rated power of the photovoltaic unit is set; g _std The irradiation power of illumination given for standard environment is usually 1kW/m2; r is R _c For a specific illumination emittance, 0.15kW/m2 is usually taken;

energy storage charge/discharge model

B(t+1)＝B(t)+P _B (t) (3-4)

3. The method for calculating the power supply reliability index of the highway self-consistent energy system according to claim 1, wherein the method comprises the following steps: a typical load day fitting method based on normal distribution is adopted to approximate a load scene of four quarters of a year, the load of 24 hours a day of a highway is considered to be approximately compliant with 24 groups of different normal distributions, and the expected value of the normal distribution in 24 hours is obtained as a traffic load demand value at the moment through sampling and is taken as a traffic load model at the demand side.

4. The method for calculating the power supply reliability index of the highway self-consistent energy system according to claim 1, wherein the method comprises the following steps: the reliability parameter model of the element comprises:

Reliability model of energy storage device

Because the energy storage system adopts a storage battery two-stage structure, the energy storage system comprises four parts: storage battery pack, DC/DC converter, DC/AC inverter and grid-connected filter, and fault rate lambda of each of the four parts of the energy storage device is obtained _C1 、λ _C2 、λ _C3 、λ _C4 And repair rate mu _c1 、μ _c2 、μ _c3 Sum mu _c4 And then the state space method is utilized, and the equivalent failure rate formula lambda is utilized _c ＝λ _c1 +λ _c2 +λ _c3 +λ _c4 And equivalent repair rate formula

Obtaining the equivalent failure rate lambda of the whole energy storage device _c (times/year) and equivalent repair rate mu _c (times/hour);

reliability model of wind turbine generator

Because the wind turbine adopts a two-stage structure, the constituent elements have four parts: wind driven generator, AC/DC rectifier, DC/AC inverter and filter, and by obtaining respective failure rate lambda of four parts of wind turbine generator _F1 、λ _F2 、λ _F3 、λ _F4 Repair rates were μ, respectively _F1 、μ _F2 、μ _F3 Sum mu _F4 And then the state space method is utilized, and the equivalent failure rate formula lambda is utilized _F ＝λ _F1 +λ _F2 +λ _F3 +λ _F4 And equivalent repair rate formula

Obtaining the equivalent failure rate lambda of the whole wind turbine generator _F (times/year) and equivalent repair rate mu _F (times/hour);

reliability model of photovoltaic unit

Due to the photovoltaic unitsAdopts a double-stage structure, and the component elements have four parts: photovoltaic arrays, DC/DC converters, DC/AC inverters, and filters. By obtaining the respective failure rate lambda of the four parts of the photovoltaic unit _G1 、λ _G2 、λ _G3 、λ _G4 Repair rates were μ, respectively _G1 、μ _G2 、μ _G3 Sum mu _G4 And then the state space method is utilized, and the equivalent failure rate formula lambda is utilized _G ＝λ _G1 +λ _G2 +λ _G3 +λ _G4 And equivalent repair rate formula

Obtaining the integral equivalent failure rate lambda of the photovoltaic unit _G (times/year) and equivalent repair rate mu _G (times/hour).

5. The method for calculating the power supply reliability index of the highway self-consistent energy system according to claim 1, wherein the method comprises the following steps: the step 4 further comprises the following steps:

(1) constructing and establishing a power model of a supply side and a demand side of the highway self-consistent energy system;

(2) establishing a reliability parameter model of each component part of the expressway self-consistent energy system;

(3) simulation is carried out: setting the total years n=1000, the simulation start years mt=1, ht=0 of the simulation to represent the simulation starting from 0 hours up to 8760 hours;

(4) determining a failed element by the minimum normal operation time T1;

(5) determining a repair time T3 of the failed component with the failed running time T2+ and the failed equipment;

(6) in the HT-HT+T1+T2+T3 time period, carrying out power balance analysis on the system;

(7) advancing the simulation time to ht=ht+t ₁ +T ₂ +T ₃ If HT<8760, returning to step (3); otherwise, executing mt=mt+1; let T ₁ ＝min(T _i ) T is then ₁ The corresponding power supply i fails. T (T) _μj ＝-(lnx _j )/μ _j ，μ _j For repairing rate, x of faulty power supply element j _j To conform to 0-1 uniformly distributed random numbers, T _μj Is the sum of the repair time T3 of the equipment with the fault running time T2+ fault, T _i For each element i normal run time.

(8) And when MT is less than or equal to N, reading 7680h data of a second group of wind-light load, returning to the step (3), and performing fault simulation of the next year.

6. A highway self-consistent energy system is characterized in that: the system comprises a non-volatile storage medium comprising a stored program, wherein the program when run controls a device in which the non-volatile storage medium is located to perform the method of claim 1.

7. An electronic device comprising a processor and a memory; the memory has stored therein computer readable instructions for execution by the processor, wherein the computer readable instructions when executed perform the method of claim 1.