Power generation system reliability evaluation method based on unit aging failure

Abstract

The invention relates to the field of power system reliability evaluation methods, in particular to a power generation system reliability evaluation method based on unit aging failure. A generator set three-state reliability model comprehensively considering an aging failure mode and a random failure mode is constructed and used for describing the time sequence state transition process of a generator set in a system under the coupling influence of aging failure and random failure; secondly, providing a system time sequence state transition sequence generation method based on a unit three-state reliability model and a sequential MCS technology; and finally, the generated system states are analyzed one by one through the optimal load reduction model, and then system probability, expectation and frequency reliability indexes are obtained. According to the invention, accurate and quantitative analysis of the reliability of the power generation system can be realized.

Description

Technical Field

[0001] The present invention relates to the field of power system reliability evaluation methods, and in particular to a power generation system reliability evaluation method based on unit aging failure. Background Art

[0002] The power generation system is an important part of the power system. Accurately evaluating the reliability level of the power generation system (i.e., evaluating the ability of the power generation system to meet power demand) can provide accurate decision-making guidance for power generation companies to carry out power supply security work. Some equipment in the power generation system has a long service life. For example, the service life of some equipment is close to 20 years, which makes the aging problem of the unit increasingly prominent. The power outage caused by aging failure will pose a serious threat to the reliability of the power generation system. The reliability assessment of the power generation system with aging and failure of the generator set is very important.

[0003] At present, the reliability assessment methods for power systems against unit aging failure include: reliability assessment methods for power generation systems based on state enumeration or non-sequential Monte Carlo sampling (MCS) method, and reliability assessment methods for power generation systems based on sequential MCS technology.

[0004] However, the existing power system reliability assessment methods have the following problems:

[0005] Reliability assessment based on state enumeration or non-sequential Monte Carlo sampling methods ignores the temporal correlation of the operating status of the generator set and the power generation system, so only probabilistic and expected reliability indicators can be obtained. In addition, such methods cannot consider the impact of strategies with temporal characteristics such as planned maintenance of units, which limits their application in engineering.

[0006] Reliability assessment based on the sequential MCS method usually uses a two-state reliability model of the generator set. First, the model assumes that the unit that has suffered aging failure can be restored to the pre-failure state through repair operations, which is inconsistent with the irreparable characteristics of unit aging failure; secondly, the model obscures the impact of the unit's random failure mode on its state transfer process; therefore, there are inevitable deviations in the reliability assessment results. Summary of the invention

[0007] The present invention aims to provide a reliability assessment method for a power generation system based on unit aging failure, so as to solve the problem of deviation in assessment results in existing reliability assessment methods.

[0008] The reliability assessment method of the power generation system based on unit aging failure in this scheme includes:

[0009] Step 1, construct a three-state reliability model of the generator set that comprehensively considers random failure and aging failure;

[0010] Step 2: Based on the historical data records of the generator sets accumulated by the power generation enterprise, a refined modeling method for the state duration in the three-state reliability model of the generator set in step 1 is used. The state duration includes the duration TTF from the unit entering the normal operating state to the first occurrence of random failure. ran , the duration from the occurrence of random failure to the completion of maintenance TTR ran , the duration TTF from the unit entering normal operation to the first occurrence of aging failure aging , the duration from the aging failure of the unit to the completion of the replacement operation TFR aging ;

[0011] Step 3: Based on the three-state reliability model of the generator set established in step 1, the component state duration sampling method in the sequential MCS method is used to generate the generator set state transition sequence considering aging failure, and the state transition sequence of each generator set in the power generation system is combined with the system load timing curve to obtain the timing transition sequence of the power generation system state;

[0012] Step 4, synchronously generate the state transition sequence of each unit in the power generation system, combine them and further split the system state with multiple load levels, and generate the state transition sequence of the entire power generation system considering the load timing curve;

[0013] Step 5, analyzing each system state included in the evaluation period based on the optimal load reduction model;

[0014] Step 6, based on the results obtained in step 5, comprehensively quantify the reliability level of the power generation system with aging and failure of computer groups from three aspects: power outage probability, expectation and frequency.

[0015] The beneficial effects of this program are:

[0016] A model is established to analyze the aging failure and random failure of the generator set, describe the timing state transfer process under the influence of the two failure couplings, generate the system timing state transfer sequence, perform the optimal solution, and obtain the evaluation indicators of the power generation system under the two failure modes. This realizes the quantitative evaluation of the reliability of the power generation system considering the aging failure of the unit, which can provide a useful reference for power generation companies to formulate operation and maintenance overhaul strategies for the units.

[0017] Furthermore, the process of constructing the three-state reliability model of the generator set in step 1 includes the following sub-steps:

[0018] Sub-step 1.1, comparative analysis of random failure modes and aging failure modes of generator sets from multiple dimensions of failure causes and failure characteristics;

[0019] Sub-step 1.2, decomposing the influence of each failure mode on the timing state transition process of the generator set and characterizing it as a combination of two different state durations;

[0020] Sub-step 1.3, use different probability distributions to perform refined modeling on the random variables of each state duration.

[0021] The beneficial effects are: comparative analysis of failure modes from multiple dimensions of the unit, quantitative representation as a combination of durations, modeling based on probability distribution, quantification of the unit status under the failure model, and facilitation of accurate judgment and analysis in subsequent steps.

[0022] Further, in step 2, the state duration TTF ran The exponential distribution is used for modeling, and its cumulative distribution function is expressed as:

[0023]

[0024] In the formula, λ ran is the average random failure rate of the generator set obtained based on historical failure data statistics.

[0025] The beneficial effect is that the duration of random failures caused by unexpected factors is quantified through exponential distribution, making the modeling results more reliable.

[0026] Furthermore, in step 2, the state duration TTR ran The modeling of is carried out using log-normal distribution, and the cumulative distribution function of log-normal distribution is expressed as:

[0027]

[0028] In the formula, μ ran and σ ran They represent the mean and standard deviation parameters of the general normal distribution, and their specific values ​​are obtained by fitting the historical failure repair records of the generator set; erf(·) is the error function, which is defined as:

[0029]

[0030] The beneficial effect is that due to factors such as maintenance personnel efficiency, weather and environmental conditions, the duration TTR may be ran There are certain fluctuations, and modeling with log-normal distribution can accurately quantify the situation caused by uncertain factors.

[0031] Further, in step 2, the state duration TTF aging The modeling is carried out using the conditional probability formula in statistics, and the cumulative distribution function of the conditional probability formula is expressed as:

[0032]

[0033] According to the analysis of historical statistical data, the cumulative aging failure distribution function F of the generator set l (t) is modeled by the three-parameter Weibull distribution, which is specifically expressed as:

[0034]

[0035] In the formula, α n and η n are the Weibull scale parameter and threshold parameter corresponding to the generator set; β is the Weibull shape parameter, α n , η n The three parameters , β are obtained by performing least square fitting on the historical failure data of the unit.

[0036] The beneficial effect is: modeling duration TTF with conditional probability formula aging , which can fully consider the actual aging changes of the unit.

[0037] Furthermore, in step 2, the state duration TFR aging Includes the state duration corresponding to the first stage states required for order creation and unit production and the duration of the state corresponding to the second stage state required for unit transportation and installation The state duration TFR aging The modeling uses log-normal distribution to independently calculate the duration of these two periods, which are expressed as:

[0038]

[0039] In the formula, and They represent the mean and standard deviation parameters of the general normal distribution function corresponding to the state duration of the first stage state respectively; and They represent the mean and standard deviation parameters of the general normal distribution function corresponding to the state duration of the second stage state respectively.

[0040] The beneficial effect is that the duration from aging failure to replacement can be accurately characterized by modeling the two stages separately.

[0041] Further, the step 3 includes the following sub-steps:

[0042] Sub-step 3.1, for any unit in the power generation system, first determine the parameters in the three-state reliability model of the unit, including the simulation period D total, the current age of the generator set T ser ;

[0043] Sub-step 3.2, initialization Assume that the unit is in operation at the initial time, set the simulated time and assign 0 to TL sim , component operation status identification variable I state Assign a value of 1, and the aging failure occurs to the indicator variable I occur Assign a value of 0; according to the current age of the crew T ser , according to the state duration TTF aging Update the cumulative distribution function formula F Tser (TTF aging );

[0044] Sub-step 3.3, determine the duration d of the unit from the current moment to the next state transition according to the preset sub-steps sim ;

[0045] Sub-step 3.4, reset the aging failure occurrence indicator variable I occur Assign a value of 0; and determine whether the preset simulation time has been reached. If not, return to sub-step 3.3. If the preset simulation time has been reached, obtain its state transition process and state duration in time period D according to the component state transition process and state duration. total The state transition sequence within.

[0046] The beneficial effect is that the state transfer sequence of each generator set in the power generation system is combined with the system load timing curve to obtain the timing transfer sequence of the power generation system state, which can dynamically analyze the changing state of the aging situation over time.

[0047] Further, in the sub-step 3.3, the preset sub-step includes:

[0048] Sub-step 3.3.1, if I state =1, indicating that the unit is in normal operation at the current moment. ran Cumulative distribution function and state duration TTF aging The cumulative distribution functions of TTF ran and TTF aging Perform random sampling and determine the next component failure type; where:

[0049]

[0050] If TTF ran_sam >TTF aging_sam , indicating that the next failure mode is random failure, update d sim TTF aging_sam , Update I state is 0, Ioccur is 1; otherwise update d sim TTF ran_sam , Update I state 0I state ←0;

[0051] Sub-step 3.3.2, if I state =0, indicating that the unit is in a failure shutdown state at the current moment. occur =0, then update I state is 1, and according to the state duration TTR ran The cumulative distribution function update d sim TTR ran_sam ,in:

[0052]

[0053] Where Φ(·) is the standard normal distribution;

[0054] Sub-step 3.3.3, if I state =0, indicating that the unit is in a failure shutdown state at the current moment. occur =1, then update I state is 1, and according to the state duration TFR aging The cumulative distribution function update d sim For TFR aging_sam ,in:

[0055]

[0056] Sub-step 3.3.4, according to the identification variable I state Determine the unit in the period [TL sim ,TL sim +d sim ] and updates the simulated time TL sim For TL sim +d sim ; Then, update the component age T according to the following formula ser :

[0057]

[0058] The beneficial effect is that the calculation of the duration between different states can accurately obtain the corresponding timing.

[0059] Furthermore, in step 5, the objective function is to minimize the total load reduction of the power generation system. Based on the DC power flow calculation model, for a given system state s, the system component state vector and the system load level are defined as v and v respectively. s and L s, an optimal load reduction model for the power generation system considering unit aging failure is established, and the model is expressed as:

[0060] The objective function is the sum of the load reduction on each bus of the power generation system LC s Minimum is the objective function of the model, expressed as:

[0061]

[0062] Where P s,i,cut represents the load reduction of node i in system state s, Ψ bus Represents a collection of system nodes;

[0063] The constraints include:

[0064] The system power balance constraint is expressed as:

[0065]

[0066] Where P s,g represents the output of the g-th generator set; Ψ g Represents the set of generator sets in the system; P i,peak represents the annual peak load of node i; L s ·P i,peak represents the load demand of node i in state s;

[0067] The node power balance constraint is expressed as:

[0068]

[0069] In the formula, Ψ g,i represents the set of generators connected to node i; P s,i,in represents the injected power of node i;

[0070] The node load reduction constraint is expressed as:

[0071]

[0072] The node injection power calculation is expressed as:

[0073]

[0074] Where n is the number of system nodes; θ s,j is the voltage phase angle of node j; b s,ij represents the element in the i-th row and j-th column of the node admittance matrix B;

[0075] The DC power flow equation for the transformer / line branch is expressed as:

[0076]

[0077] In the formula, θ s,i and θ s,j Respectively represent the voltage phase angles of the first and last nodes i and j of branch r; s,r Represents the reactance of branch r; PL s,r represents the power flow of branch r; branch Represents a set of system branches;

[0078] The transformer / line branch transfer capacity constraint is expressed as:

[0079]

[0080] In the formula, Indicates the maximum transmission capacity of branch r; state identification variable υ s,r Used to describe the operating status of branch r;

[0081] The output constraint of the generator set is expressed as:

[0082]

[0083] In the formula, Indicates the maximum output of the generator set g, which is taken as the rated capacity; υ s,g Used to describe the operating status of the generator set g: When the generator set g stops operating due to random failure or aging, υ s,g The value is 0, otherwise s,g is 1;

[0084] For system state s, if LC s >0 means that the power generation system cannot fully meet the load supply in this state, and there is load reduction. At this time, the load reduction identification variable J needs to be set. s Recorded as 1; if LC s =0, then J s is 0;

[0085] In the reliability assessment process based on sequential MCS, the LC of each system state is recorded. s and J s The calculation results are used in step 6 to calculate the reliability index of the power generation system.

[0086] The beneficial effect is that the setting of constraint conditions can make the operation of the overall power generation system more economical and safer while taking aging failure into consideration.

[0087] Furthermore, in step 6, the process of quantifying the reliability level of the power generation system due to aging failure of the computer group includes the following sub-steps:

[0088] Sub-step 6.1, calculate the load reduction probability LOLP index of the power generation system, the formula is:

[0089]

[0090] In the formula, K sam Indicates the total number of evaluation cycles, which can usually be set to 5000 times; D total Represents the length of the reliability assessment cycle, usually 1 year; Θ k represents the set of all system states in the kth evaluation cycle; J k,s and d k,s They represent the load reduction identification variable and state duration corresponding to the system state s in the kth evaluation period respectively;

[0091] Sub-step 6.2, calculate the expected power shortage EENS index of the power generation system, expressed as:

[0092]

[0093] In the formula, LC k,s represents the load reduction per unit time of system state s in the kth evaluation period;

[0094] Sub-step 6.3, calculate the expected power shortage LOLF index of the power generation system, expressed as:

[0095]

[0096] In the formula, I k,s is a 0-1 flag variable. If the system state s has load shedding and its previous state does not have load shedding, then I k,s The value is 1, otherwise I k,s The value is 0.

[0097] The beneficial effect is that the calculation of various quantitative indicators can evaluate the reliability of the power generation system from multiple aspects. BRIEF DESCRIPTION OF THE DRAWINGS

[0098] Figure 1 It is a flowchart of an embodiment of a method for evaluating reliability of a power generation system based on aging failure of a unit according to the present invention;

[0099] Figure 2 It is a reliability assessment flow chart of an embodiment of a method for assessing the reliability of a power generation system based on unit aging failure of the present invention;

[0100] Figure 3 It is a time sequence state transfer process diagram of a generator set in an embodiment of a method for evaluating the reliability of a power generation system based on aging failure of a generator set of the present invention;

[0101] Figure 4A three-state reliability model of a generator set considering aging failure in an embodiment of a method for evaluating reliability of a power generation system based on aging failure of a generator set according to the present invention;

[0102] Figure 5 It is a process diagram of generating a state transfer sequence of a power generation system based on sequential MCS in an embodiment of the method for evaluating the reliability of a power generation system based on unit aging failure of the present invention. DETAILED DESCRIPTION

[0103] The following is further explained in detail through specific implementation methods.

[0104] Example

[0105] Reliability assessment methods for power generation systems based on unit aging failure, such as Figure 1 and Figure 2 As shown, including:

[0106] Step 1: construct a three-state reliability model of the generator set that comprehensively considers random failures and aging failures. The process of constructing the three-state reliability model of the generator set includes the following sub-steps:

[0107] Sub-step 1.1, compare and analyze the random failure mode and aging failure mode of the generator set from multiple dimensions of failure causes and failure characteristics. Any failure will cause the unit to transfer from the normal operation state to the failure shutdown state. In order to quantitatively characterize the differential impact of the two failure modes on the sequential state transition process of the generator set, the causes, failure rates, and failure frequencies of the two failure modes are first compared and analyzed. The specific analysis results are shown in Table 1.

[0108] Table 1 Comparison between aging failure mode and random failure mode of generator set

[0109]

[0110] It can be seen from Table 1 that the essential difference between random failure and aging failure modes is that random failure is usually repairable and has nothing to do with the unit's operating time, while aging failure is related to the unit's operating time and is irreparable.

[0111] Sub-step 1.2: From sub-step 1.1, we know that when the generator set fails due to aging, it needs to be replaced with a new one to restore the system function. Before that, the generator set will alternate between normal operation and random failure. Therefore, the schematic diagram of the timing state transition process of the generator set considering aging failure is as follows: Figure 3 The influence of each failure mode on the timing state transition process of the generator set is decomposed and characterized as a combination of two different state durations.

[0112] Sub-step 1.3: From sub-step 1.2, we can know that the timing state transition process of the generator set can be regarded as a timing combination of two operation cycles corresponding to the random failure mode and the aging failure mode. Among them, the operation cycle corresponding to the random failure mode can be decomposed into TTF ran and TTR ran Two state duration random variables; in this operation cycle, the unit first fails randomly after running for a period of time, causing it to switch from the operating state to the outage state; after the maintenance operation is completed, the unit will return to the normal operating state, and the operating time continues to accumulate. Accordingly, the operation cycle corresponding to the aging failure mode can be regarded as the state duration random variable TTF aging and TFR aging The unit will age and fail after a period of normal operation during the operation cycle, and then transfer to the shutdown state; after the replacement operation is completed, the unit will transfer back to the normal operation state. Different probability distributions are used to fine-tune the random variables of each state duration, and the three-state reliability model is constructed as follows: Figure 4 shown.

[0113] Step 2: Based on the historical data records of the generator sets accumulated by the power generation enterprise, a refined modeling method for the state duration in the three-state reliability model of the generator set in step 1 is used. The state duration includes the duration TTF from the unit entering the normal operating state to the first occurrence of random failure. ran , the duration from the occurrence of random failure to the completion of maintenance TTR ran , the duration TTF from the unit entering normal operation to the first occurrence of aging failure aging , the duration from the aging failure of the unit to the completion of the replacement operation TFR aging .

[0114] Random failure of generator sets is usually caused by unexpected factors, and the state duration TTF ran The exponential distribution is used for modeling, and its cumulative distribution function is expressed as:

[0115]

[0116] In the formula, λ ran is the average random failure rate of the generator set obtained based on historical failure data statistics.

[0117] Maintenance time TFR aging Due to factors such as maintenance personnel efficiency, weather and environmental conditions, there may be certain fluctuations. Status Duration TTR ran The modeling of is carried out using log-normal distribution, and the cumulative distribution function of log-normal distribution is expressed as:

[0118]

[0119] In the formula, μ ran and σ ran They represent the mean and standard deviation parameters of the general normal distribution, and their specific values ​​are obtained by fitting the historical failure repair records of the generator set; erf(·) is the error function, which is defined as:

[0120]

[0121] Different from TTF aging and TFR agingn , state duration TTF aging TTF related to the running time of the generator set aging The cumulative distribution function of will change dynamically with the running time of the generator set. Assume that the cumulative aging failure distribution function of the generator set is F l (t), then its running time is T ser Under the condition of aging The modeling is carried out using the conditional probability formula in statistics, and the cumulative distribution function of the conditional probability formula is expressed as:

[0122]

[0123] According to the analysis of historical statistical data, the cumulative aging failure distribution function F of the generator set l (t) is modeled by the three-parameter Weibull distribution, which is specifically expressed as:

[0124]

[0125] In the formula, α n and η n are the Weibull scale parameter and threshold parameter corresponding to the generator set; β is the Weibull shape parameter, α n , η n The three parameters , β are obtained by performing least square fitting on the historical failure data of the unit.

[0126] Generator set replacement time TFR aging It will be affected by many factors such as the manufacturer's manufacturing progress, transportation and on-site installation conditions. Therefore, the state duration TFR aging Includes the state duration corresponding to the first stage states required for order creation and unit production and the duration of the state corresponding to the second stage state required for unit transportation and installation Total replacement time TFR aging It can be expressed as The state duration TFRaging The modeling uses log-normal distribution to independently calculate the duration of these two periods, which are expressed as:

[0127]

[0128] In the formula, and They represent the mean and standard deviation parameters of the general normal distribution function corresponding to the state duration of the first stage state respectively; and They represent the mean and standard deviation parameters of the general normal distribution function corresponding to the state duration of the second stage state respectively.

[0129] Step 3: Based on the three-state reliability model of the generator set established in step 2, the component state duration sampling method in the sequential MCS method is used to generate the generator set state transition sequence considering aging failure, and the state transition sequence of each generator set in the power generation system is combined with the system load timing curve to obtain the timing transition sequence of the power generation system state. Specifically, it includes the following sub-steps:

[0130] Sub-step 3.1, for any unit in the power generation system, first determine the parameters in the three-state reliability model of the unit, including the simulation period D total , the current age of the generator set T ser ;

[0131] Sub-step 3.2, initialization assumes that the unit is in operation at the initial moment and sets the simulated time TL sim ←0, the arrows in this article represent assignments, i.e., to TL sim Assign 0, component operation status identification variable I state ←1, that is, to I state Assign value 1, aging failure occurs to identify variable I occur ←0, that is, to I occur Assign a value of 0; according to the crew's current age T ser , according to the state duration TTF aging Update the cumulative distribution function formula F Tser (TTF aging );

[0132] Sub-step 3.3, determine the duration d of the unit from the current moment to the next state transition according to the preset sub-steps sim , the preset sub-steps include:

[0133] Sub-step 3.3.1, if I state =1, indicating that the unit is in normal operation at the current moment. ranCumulative distribution function and state duration TTF aging The cumulative distribution functions of TTF ran and TTF aging Perform random sampling and determine the next component failure type; where:

[0134]

[0135] If TTF ran_sam >TTF aging_sam , indicating that the next failure mode is random failure, update d sim ←TTF aging_sam , that is, to d sim Assign TTF aging_sam , Update I state ←0,I occur ←1, that is, to I state Assign 0 to I occur Assign 1; otherwise update d sim ←TTF ran_sam , that is, to d sim Assign TTF ran_sam , Update I state ←0, that is, to I state Assign 0;

[0136] Sub-step 3.3.2, if I state =0, indicating that the unit is in a failure shutdown state at the current moment. occur =0, then update I state ←1, that is, to I state Assign a value of 1 and set the state duration TTR ran The cumulative distribution function update d sim ←TTR ran_sam , that is, to d sim Assign TTR ran_sam ,in:

[0137]

[0138] Where Φ(·) is the standard normal distribution;

[0139] Sub-step 3.3.3, if I state =0, indicating that the unit is in a failure shutdown state at the current moment. occur =1, then update I state ←1, that is, to I state Assign a value of 1 and set the state duration TFR aging The cumulative distribution function update d sim ←TFR aging_sam , that is, to d sim Assign TFRaging_sam ,in:

[0140]

[0141] Sub-step 3.3.4, according to the identification variable I state Determine the unit in the period [TL sim ,TL sim +d sim ] and updates the simulated time TL sim ←TL sim +d sim ; Then, update the component age T according to the following formula ser :

[0142]

[0143] In the formulas in this article, unexplained parameters represent intermediate operational variables and have no actual meaning;

[0144] Sub-step 3.4, reset the aging failure occurrence indicator variable I occur ←0; and determine whether the preset simulation time has been reached. If not, return to sub-step 3.3. If the preset simulation time has been reached, obtain its state transition process and state duration in time period D according to the component state transition process and state duration. total The state transition sequence within.

[0145] Step 4, synchronously generate the state transition sequence of each unit in the power generation system, combine them and further split the system state with multiple load levels to generate the state transition sequence of the entire power generation system considering the load timing curve. Figure 5 As shown in FIG. 1 , the generation process of the system state transition sequence S2 containing three components (C1-C3) and three load levels (Ls: L1-L3) is shown. In which, indexes 1-15 represent 15 independent system states, and each system is connected to its adjacent states in chronological order.

[0146] Step 5: Analyze each system state included in the evaluation cycle based on the optimal load reduction model. Specifically, in the process of evaluating the reliability of the power generation system, it is assumed that the power system lines, transformers and other equipment are completely reliable. Taking the minimum total load reduction of the power generation system as the objective function, based on the DC power flow calculation model, for a given system state s, define its system component state vector and system load level as v respectively. s and L s , an optimal load reduction model for the power generation system considering unit aging failure is established, and the model is expressed as:

[0147] The objective function is the sum of the load reduction on each bus (node) of the power generation system LCs Minimum is the objective function of the model, expressed as:

[0148]

[0149] Where P s,i,cut represents the load reduction of node i in system state s, Ψ bus Represents a collection of system nodes;

[0150] The constraints include:

[0151] The system power balance constraint is expressed as:

[0152]

[0153] Where P s,g represents the output of the g-th generator set; Ψ g Represents the set of generator sets in the system; P i,peak represents the annual peak load of node i; L s ·P i,peak represents the load demand of node i in state s.

[0154] The node power balance constraint is expressed as:

[0155]

[0156] In the formula, Ψ g,i represents the set of generators connected to node i; P s,i,in represents the injected power of node i.

[0157] The node load reduction constraint is expressed as:

[0158]

[0159] The node injection power calculation is expressed as:

[0160]

[0161] Where n is the number of system nodes; θ s,j is the voltage phase angle of node j; b s,ij Represents the element in the i-th row and j-th column of the node admittance matrix B.

[0162] The DC power flow equation for the transformer / line branch is expressed as:

[0163]

[0164] In the formula, θ s,i and θ s,j Respectively represent the voltage phase angles of the first and last nodes i and j of branch r; s,rRepresents the reactance of branch r; PL s,r represents the power flow of branch r; branch Indicates the system branch set.

[0165] The transformer / line branch transfer capacity constraint is expressed as:

[0166]

[0167] In the formula, Indicates the maximum transmission capacity of branch r; state identification variable υ s,r Used to describe the operating status of branch r. Since it is completely reliable, the value is 1.

[0168] The output constraint of the generator set is expressed as:

[0169]

[0170] In the formula, Indicates the maximum output of the generator set g, which is taken as the rated capacity; υ s,g Used to describe the operating status of the generator set g: When the generator set g stops operating due to random failure or aging, υ s,g The value is 0, otherwise s,g is 1.

[0171] For system state s, if LC s >0 means that the power generation system cannot fully meet the load supply in this state, and there is load reduction. At this time, the load reduction identification variable J needs to be set. s Recorded as 1; if LC s =0, then J s is 0.

[0172] In the reliability assessment process based on sequential MCS, the LC of each system state is recorded. s and J s The calculation results are used in step 6 to calculate the reliability index of the power generation system.

[0173] Step 6: Based on the result obtained in step 5, the reliability level of the power generation system due to the aging failure of the computer group is comprehensively quantified from the three aspects of power outage probability, expectation and frequency. Specifically, the process of quantifying the reliability level of the power generation system due to the aging failure of the computer group includes the following sub-steps:

[0174] Sub-step 6.1, calculate the load reduction probability LOLP index of the power generation system, the formula is:

[0175]

[0176] In the formula, K sam Indicates the total number of evaluation cycles, usually set to 5000 times; Dtotal Represents the length of the reliability assessment cycle, usually 1 year; Θ k represents the set of all system states in the kth evaluation cycle; J k,s and d k,s They represent the load reduction identification variable and state duration corresponding to the system state s in the kth evaluation period respectively;

[0177] Sub-step 6.2, calculate the expected power shortage EENS index of the power generation system, expressed as:

[0178]

[0179] In the formula, LC k,s represents the load reduction per unit time of system state s in the kth evaluation period;

[0180] Sub-step 6.3, calculate the expected power shortage LOLF index of the power generation system, expressed as:

[0181]

[0182] In the formula, I k,s is a 0-1 flag variable. If the system state s has load shedding and its previous state does not have load shedding, then I k,s The value is 1, otherwise I k,s The value is 0.

[0183] In this embodiment, firstly, the proposed three-state reliability model can accurately simulate the sequential state transition process and failure behavior of the generator set under the combined influence of the aging failure mode and the random failure mode. Secondly, the proposed reliability assessment method based on sequential MCS can accurately quantify the reliability level of the power generation system considering the aging failure of the unit from three dimensions: power outage probability, expectation and frequency. Finally, the proposed assessment method is also applicable to operation and maintenance strategy decisions with sequential characteristics, such as planned maintenance analysis of generator sets, and is more suitable for engineering applications than traditional methods.

[0184] Combined with the aging failure and random failure of the generator set, the model is established for analysis, the time sequence state transfer process under the influence of the two failure couplings is described, the system time sequence state transfer sequence is generated, the optimal solution is obtained, and the evaluation index of the power generation system under the two failure modes is obtained, which realizes the quantitative evaluation of the reliability of the power generation system considering the aging failure of the unit. Based on this scheme, the operating status of the power generation system can be fully understood, and a guiding basis can be provided for making reasonable operation and planning decisions and preventing major power outages in the system. This scheme can provide a useful reference for power generation companies to formulate operation and maintenance strategies for the units.

[0185] The above is only an embodiment of the present invention, and the common knowledge such as the known specific structure and characteristics in the scheme is not described in detail here. It should be pointed out that for those skilled in the art, several deformations and improvements can be made without departing from the structure of the present invention, which should also be regarded as the protection scope of the present invention, and these will not affect the effect of the implementation of the present invention and the practicality of the patent. The scope of protection required by this application shall be based on the content of its claims, and the specific implementation methods and other records in the specification can be used to interpret the content of the claims.

Claims

1. A reliability assessment method for a power generation system based on unit aging failure, characterized in that: include: Step 1, construct a three-state reliability model of the generator set that comprehensively considers random failure and aging failure; Step 2: Based on the historical data records of the generator sets accumulated by the power generation enterprise, a refined modeling method for the state duration in the three-state reliability model of the generator set in step 1 is used. The state duration includes the duration TTF from the unit entering the normal operating state to the first occurrence of random failure. ran , the duration from the occurrence of random failure to the completion of maintenance TTR ran , the duration TTF from the unit entering normal operation to the first occurrence of aging failure aging , the duration from the aging failure of the unit to the completion of the replacement operation TFR aging ; Step 3: Based on the three-state reliability model of the generator set established in step 2, the component state duration sampling method in the sequential MCS method is used to generate the generator set state transition sequence considering aging failure, and the state transition sequence of each generator set in the power generation system is combined with the system load timing curve to obtain the timing transition sequence of the power generation system state; Step 4, synchronously generate the state transition sequence of each unit in the power generation system, combine them and further split the system state with multiple load levels, and generate the state transition sequence of the entire power generation system considering the load timing curve; Step 5, analyzing each system state included in the evaluation period based on the optimal load reduction model; Step 6, based on the results obtained in step 5, comprehensively quantify the reliability level of the power generation system with aging and failure of computer groups from three aspects: power outage probability, expectation and frequency.

2. The method for evaluating the reliability of a power generation system based on unit aging failure according to claim 1 is characterized in that: The process of constructing the three-state reliability model of the generator set in step 1 includes the following sub-steps: Sub-step 1.1, comparative analysis of random failure modes and aging failure modes of generator sets from multiple dimensions of failure causes and failure characteristics; Sub-step 1.2, decomposing the influence of each failure mode on the timing state transition process of the generator set and characterizing it as a combination of two different state durations; Sub-step 1.3, use different probability distributions to perform refined modeling on the random variables of each state duration.

3. The method for evaluating the reliability of a power generation system based on unit aging failure according to claim 1 is characterized in that: In step 2, the state duration TTF ran The exponential distribution is used for modeling, and its cumulative distribution function is expressed as: In the formula, λ ran is the average random failure rate of the generator set obtained based on historical failure data statistics.

4. The method for evaluating the reliability of a power generation system based on unit aging failure according to claim 1 is characterized in that: In step 2, the state duration TTR ran The modeling of is carried out using log-normal distribution, and the cumulative distribution function of log-normal distribution is expressed as: In the formula, μ ran and σ ran They represent the mean and standard deviation parameters of the general normal distribution, and their specific values ​​are obtained by fitting the historical failure repair records of the generator set; erf(·) is the error function, which is defined as:

5. The method for evaluating the reliability of a power generation system based on unit aging failure according to claim 1 is characterized in that: In step 2, the state duration TTF aging The modeling is carried out using the conditional probability formula in statistics, and the cumulative distribution function of the conditional probability formula is expressed as: According to the analysis of historical statistical data, the cumulative aging failure distribution function F of the generator set l (t) is modeled by the three-parameter Weibull distribution, which is specifically expressed as: In the formula, α n and η n are the Weibull scale parameter and threshold parameter corresponding to the generator set; β is the Weibull shape parameter, α n , η n The three parameters , β are obtained by performing least square fitting on the historical failure data of the unit.

6. The method for evaluating the reliability of a power generation system based on unit aging failure according to claim 1 is characterized in that: In step 2, the state duration TFR aging Includes the state duration corresponding to the first stage states required for order creation and unit production and the duration of the state corresponding to the second stage state required for unit transportation and installation The state duration TFR aging The modeling uses log-normal distribution to independently calculate the duration of these two periods, which are expressed as: In the formula, and They represent the mean and standard deviation parameters of the general normal distribution function corresponding to the state duration of the first stage state respectively; and They represent the mean and standard deviation parameters of the general normal distribution function corresponding to the state duration of the second stage state respectively.

7. The method for evaluating the reliability of a power generation system based on unit aging failure according to claim 1 is characterized in that: The step 3 includes the following sub-steps: Sub-step 3.1, for any unit in the power generation system, first determine the parameters in the three-state reliability model of the unit, including the simulation period D total , the current age of the generator set T ser ; Sub-step 3.2, initialization Assume that the unit is in operation at the initial time, set the simulated time and assign 0 to TL sim , component operation status identification variable I state Assign a value of 1, and the aging failure occurs to the indicator variable I occur Assign a value of 0; according to the current age of the crew T ser , update F according to the cumulative distribution function formula of the state duration TTFaging Tser (TTF aging ); Sub-step 3.3, determine the duration d of the unit from the current moment to the next state transition according to the preset sub-steps sim ; Sub-step 3.4, reset the aging failure occurrence indicator variable I occur Assign a value of 0; and determine whether the preset simulation time has been reached. If not, return to sub-step 3.

3. If the preset simulation time has been reached, obtain its state transition process and state duration in time period D according to the component state transition process and state duration. total The state transition sequence within.

8. The method for evaluating the reliability of a power generation system based on unit aging failure according to claim 7 is characterized in that: In the sub-step 3.3, the preset sub-step includes: Sub-step 3.3.1, if I state =1, indicating that the unit is in normal operation at the current moment. ran Cumulative distribution function and state duration TTF aging The cumulative distribution functions of TTF ran and TTF aging Perform random sampling and determine the next component failure type; where: if TTF ran_sam >TTF aging_sam , indicating that the next failure mode is random failure, update d sim TTF aging_sam , Update I state is 0, I occur is 1; otherwise update d sim TTF ran_sam , Update I state is 0; Sub-step 3.3.2, if I state =0, indicating that the unit is in a failure shutdown state at the current moment. occur =0, then update I state is 1, and according to the state duration TTR ran The cumulative distribution function update d sim TTR ran_sam ,in: Where Φ(·) is the standard normal distribution; Sub-step 3.3.3, if I state =0, indicating that the unit is in a failure shutdown state at the current moment. occur =1, then update I state is 1, and according to the state duration TFR aging The cumulative distribution function update d sim For TFR aging_sam ,in: Sub-step 3.3.4, according to the identification variable I state Determine the unit in the period [TL sim ,TL sim +d sim ] and updates the simulated time TL sim For TL sim +d sim ; Then, update the component age T according to the following formula ser :

9. The method for evaluating the reliability of a power generation system based on unit aging failure according to claim 8 is characterized in that: In step 5, the objective function is to minimize the total load reduction of the power generation system. Based on the DC power flow calculation model, for a given system state s, the system component state vector and the system load level are defined as v s and L s , an optimal load reduction model for the power generation system considering unit aging failure is established, and the model is expressed as: The objective function is the sum of the load reduction on each bus of the power generation system LC s Minimum is the objective function of the model, expressed as: Where P s,i,cut represents the load reduction of node i in system state s, Ψ bus Represents a collection of system nodes; The constraints include: The system power balance constraint is expressed as: Where P s,g represents the output of the g-th generator set; Ψ g Represents the set of generator sets in the system; P i,peak represents the annual peak load of node i; L s ·P i,peak represents the load demand of node i in state s; The node power balance constraint is expressed as: In the formula, Ψ g,i represents the set of generators connected to node i; P s,i,in represents the injected power of node i; The node load reduction constraint is expressed as: The node injection power calculation is expressed as: Where n is the number of system nodes; θ s,j is the voltage phase angle of node j; b s,ij represents the element in the i-th row and j-th column of the node admittance matrix B; The DC power flow equation for the transformer / line branch is expressed as: In the formula, θ s,i and θ s,j Respectively represent the voltage phase angles of the first and last nodes i and j of branch r; s,r Represents the reactance of branch r; PL s,r represents the power flow of branch r; branch Represents a set of system branches; The transformer / line branch transfer capacity constraint is expressed as: In the formula, Indicates the maximum transmission capacity of branch r; state identification variable υ s,r Used to describe the operating status of branch r; The output constraint of the generator set is expressed as: In the formula, Indicates the maximum output of the generator set g, which is taken as the rated capacity; υ s,g Used to describe the operating status of the generator set g: When the generator set g stops operating due to random failure or aging, υ s,g The value is 0, otherwise s,g is 1; For system state s, if LC s >0 means that the power generation system cannot fully meet the load supply in this state, and there is load reduction. At this time, the load reduction identification variable J needs to be set. s Recorded as 1; if LC s =0, then J s is 0; In the reliability assessment process based on sequential MCS, the LC of each system state is recorded. s and J s The calculation results are used in step 6 to calculate the reliability index of the power generation system.

10. The method for evaluating the reliability of a power generation system based on unit aging failure according to claim 9, characterized in that: In step 6, the process of quantifying the reliability level of the power generation system due to aging failure of the computer group includes the following sub-steps: Sub-step 6.1, calculate the load reduction probability LOLP index of the power generation system, the formula is: In the formula, K sam Indicates the total number of evaluation cycles, which can usually be set to 5000 times; D total Represents the length of the reliability assessment cycle, usually 1 year; Θ k represents the set of all system states in the kth evaluation cycle; J k,s and d k,s They represent the load reduction identification variable and state duration corresponding to the system state s in the kth evaluation period respectively; Sub-step 6.2, calculate the expected power shortage EENS index of the power generation system, expressed as: In the formula, LC k,s represents the load reduction per unit time of system state s in the kth evaluation period; Sub-step 6.3, calculate the expected power shortage LOLF index of the power generation system, expressed as: In the formula, I k,s is a 0-1 flag variable. If the system state s has load shedding and its previous state does not have load shedding, then I k,s The value is 1, otherwise I k,s The value is 0.

Images