Offshore charging station reliability improvement method considering time-varying fault and new energy randomness

Abstract

The invention discloses an offshore charging station reliability improvement method considering time-varying fault and new energy randomness, and belongs to the technical field of charging station reliability evaluation, and the method comprises the steps: S1, constructing an evaluation index for evaluating the power supply reliability of an electric ship charging station; s2, constructing an electric ship charging station reliability evaluation model; s3, based on Monte Carlo sampling, generating a fault state time sequence, gradually updating the component state in the time sequence until the repair is completed, and further obtaining a reliability evaluation result of the offshore charging station; s4, charging station redundancy configuration is carried out based on reliability requirements and cost control, and the evaluation index of S1 and the reliability evaluation result of S3 are fitted; s5, based on the reliability requirement, establishing a reliability improvement model; according to the marine charging station reliability improvement method considering the time-varying fault and the new energy randomness provided by the invention, the reliability and economical efficiency of the system are remarkably improved by constructing a multi-dimensional reliability index model and a data-driven redundancy configuration optimization mechanism.

Description

Technical Field

[0001] The present invention relates to the technical field of reliability assessment of charging stations, and particularly to a method for improving the reliability of offshore charging stations considering time-varying faults and the randomness of new energy. Background Art

[0002] With the rapid development of offshore electric transportation, offshore charging stations, as the core infrastructure to support the endurance of electric ships, their power supply reliability has become the key bottleneck restricting the transformation of marine clean energy. Compared with the mature onshore charging system, offshore charging stations face the dual challenges of dynamic marine environment and the vulnerability of complex systems, leading to the structural dilemma of power supply reliability.

[0003] At the energy supply architecture level, traditional onshore charging stations rely on high-reliability land distribution networks, while offshore charging stations need to rely on offshore renewable energy with significant intermittency and volatility. The strong coupling effect of meteorological conditions on wind and light resources leads to the spatio-temporal heterogeneity of energy supply stability, forming a power balance problem with the superposition of uncertainties on both the "source-load" sides. In terms of system composition, as a multi-device coupling system integrating new energy generation, energy storage buffering, and charging load, offshore charging stations have a non-linear cumulative effect on the reliability risk caused by the time-varying failure rate of their components.

[0004] Existing research mainly builds evaluation models based on the reliability theory of distribution systems for single uncertainties such as renewable energy fluctuations and equipment failures. However, the reliability mechanism of offshore charging stations presents multi-dimensional complexity characteristics: in the dimension of uncertainty sources, in addition to endogenous faults of equipment (such as the capacity attenuation of energy storage systems and the aging of wind turbine blades), exogenous faults caused by the variability of marine meteorology (such as the sudden power drop in extreme weather and continuous windless and lightless conditions) also need to be considered; in the time scale dimension, the time-series charge and discharge characteristics of energy storage systems and the dynamic evolution of failure rates form a cross-time coupling effect; in the spatial distribution dimension, the geographical differences of wind and light resources in sea areas lead to the need to consider regional characteristics in the assessment of power supply reliability. The current methods are difficult to meet the decision-making requirements of reliability-economy collaborative optimization in engineering practice.

[0005] Therefore, aiming at the special operating environment and system architecture of offshore charging stations, it is urgent to establish a new paradigm for power supply reliability assessment that integrates multi-source uncertainty modeling, equipment state evolution analysis, and energy storage time-series characteristic characterization, realizes the methodological transition from single-fault analysis to system resilience enhancement, and provides theoretical support and technical paths for the engineering deployment of offshore charging stations.

[0006] The prior art one can refer to a reliability analysis system for a new energy electric vehicle charging pile, including a data acquisition module, an idle analysis module, a charging analysis module, a reliability analysis module and a safety warning module. By specifically analyzing the temperature and power of the new energy electric vehicle charging pile in the idle, slow charging and fast charging states, abnormal conditions of the charging pile in different working states can be accurately identified, and abnormal conditions can be discovered in time. Although the above solution can identify the faults of the device itself, it has the following defects: the reliability evaluation dimension is single, the correlation between redundant configuration and reliability is missing, it relies on empirical expansion strategies, which is likely to cause insufficient redundancy or cost waste; the dynamic risk modeling is insufficient; the optimization efficiency is low, and it is difficult to achieve the rapid economic optimization of a multi-charging station network. The prior art two can refer to a reliability evaluation method and system for the operating state of a charging pile, which can timely discover the faults and actual operating state of the charging pile, facilitating the adoption of corresponding treatment measures to improve the charging efficiency. Although it can identify abnormal device operation, it still has the following defects: the evaluation scope is limited to the efficiency of the charging pile itself, and the influence of external environmental factors on power supply reliability is not considered, resulting in the lack of system-level risk assessment; the reliability index is single; the correlation between redundant configuration and reliability is not modeled; the coverage of dynamic risk scenarios is insufficient, and the reliability evaluation accuracy in extreme scenarios is insufficient; the optimization mechanism is missing, and there is a lack of a closed-loop optimization model with cost minimization as the goal and reliability as the constraint, resulting in low resource allocation efficiency. Summary of the Invention

[0007] The object of the present invention is to provide a method for improving the reliability of a marine charging station considering time-varying faults and new energy randomness, so as to solve the problems in the above-mentioned background technology that the existing reliability evaluation system relies on a single index, the redundant design lacks a quantitative model, it is difficult to achieve the rapid economic optimization of a large-scale charging station network, and the reliability evaluation accuracy in extreme scenarios is insufficient.

[0008] To achieve the above object, the present invention provides a method for improving the reliability of a marine charging station considering time-varying faults and new energy randomness, including the following steps:

[0009] S1. Construct evaluation indexes for evaluating the power supply reliability of an electric ship charging station;

[0010] S2. Construct a reliability evaluation model for an electric ship charging station, including component reliability and system reliability;

[0011] S3. Based on Monte Carlo sampling, generate a time series of fault states, and gradually update the component states within the time series until the repair is completed, so as to obtain the reliability evaluation result of the marine charging station;

[0012] S4. Carry out redundant configuration of the charging station based on reliability requirements and cost control, and fit the evaluation indexes in S1 with the reliability evaluation result in S3;

[0013] S5. Based on the reliability requirements, establish a reliability improvement model.

[0014] Preferably, the evaluation indexes in step S1 include:

[0015] For a charging station, the power supply reliability of the charging station is measured by the expected energy not supplied (EENS) and the mean time to repair (MTTR), which are respectively expressed as:

[0016]

[0017]

[0018] Among them, is the weight of each charging station on the route, which is determined according to the installed capacity of each charging station; represents the load power of charging station i at time t; represents the charging power of the i-th charging station at time t; n is the total number of charging stations; Δt is the time interval; represents the repair time of the f-th failure of the i-th charging station; represents the total number of failures of the i-th charging station; F i The failure set of the i-th charging station;

[0019] For an electric ship, its reliability is measured by the loss of load probability (LOLP), the system average interruption duration index (SAIDI) and the system average interruption frequency index (SAIFI), which are respectively expressed as:

[0020]

[0021]

[0022]

[0023] Among them, is an indicator function. If the load demand of the i-th charging station at time t is greater than the available power, it takes the value of 1, otherwise it is 0; U i,t represents the number of users affected by the power outage at the i-th charging station at time t; α i,t is an indicator function. If the i-th charging station is operating normally at time t, it takes the value of 1, otherwise it is 0; is the total number of users of the i-th charging station.

[0024] Preferably, the component reliability evaluation model in step S2 includes:

[0025] The failure rate model of the charging equipment is modeled by the exponential distribution, and the failure distribution is expressed as:

[0026] f t cf= λe -λt (6)

[0027] where λ is the failure rate constant;

[0028] The energy storage system is modeled using the Weibull distribution, and its failure distribution is expressed as:

[0029]

[0030] where β is the shape parameter; η is the scale parameter;

[0031] The offshore wind turbine is modeled using the normal distribution, and the turbine failure distribution is expressed as;

[0032]

[0033] where σ is the standard deviation, used to describe the discreteness of the failure time; μ is the mean, representing the average failure life of the wind turbine.

[0034] Preferably, for the system reliability evaluation model in step S2, O t represents its failure state at time t. If it fails at time t, then O t = 0, otherwise it is 1. Since the failure state is related to time, the failure condition at the next moment is inferred based on the previous moment, and is expressed as:

[0035] P(O t+1 = 0|O t = 1) = f t (9)

[0036] where P(O t+1 = 0|O t = 1) represents the probability that the component still fails at time t + 1 when it is operating normally at time t; if it fails, its repair probability is p re , and is expressed as:

[0037] P(O t+1 = 1|O t = 0) = p re (10)

[0038] Assume that the repair time interval is T re , then the formulas (9) and (10) are corrected to obtain:

[0039]

[0040] Preferably, step S3 specifically includes:

[0041] S31. Conduct Monte Carlo sampling on the equipment failure state to generate a failure state time series;

[0042] S32. Obtain the sampling time After that, gradually update the component status within the time series; if the current time then component c i is marked as the fault status until the repair is completed;

[0043] S33. After obtaining the Monte Carlo sampling data, substitute the component fault status into the system operation simulation model to obtain the total output of the new energy unit and the energy storage system, and calculate its power gap.

[0044] Preferably, step S31 is specifically as follows:

[0045] First, define the system component set C = {c1, c2,... c n}, the total number of simulation times M, and the time series T = {t1, t2,... t k}, where k is the number of time steps;

[0046] For each component c i , randomly generate a set of samples based on its fault distribution; for the exponential distribution, generate the sampling time through the inverse transformation method:

[0047]

[0048] For the Weibull distribution, perform reverse sampling according to its cumulative distribution function:

[0049]

[0050] For the normal distribution, generate normal samples randomly:

[0051]

[0052] Preferably, step S33 is specifically as follows:

[0053] For each charging station i, the output of its new energy unit is corrected as:

[0054]

[0055] where j represents the number of the new energy unit in charging station i; N j represents the total number of new energy units; is the fault status of each unit; represents the output of each unit; and the remaining power of the new energy unit is:

[0056]

[0057] where L i,t represents the charging demand of this charging station at this moment; assume the discharge power of the energy storage system is then:

[0058]

[0059] Among them, is the fault state of the energy storage system; after obtaining the discharge power, its energy S is deduced i,j,t Change situation:

[0060]

[0061] Similarly, the power situation and energy change situation of the energy storage system during charging are obtained:

[0062]

[0063]

[0064] Among them, η dis represents the discharge efficiency; represents the maximum capacity of the energy storage system; η ch represents the charging efficiency;

[0065] According to the above derivation, the total output of the new energy unit and the energy storage system is obtained, and its power gap is calculated.

[0066] Preferably, in step S4, spline interpolation is used as the fitting method, and the fitting process is as follows:

[0067] For charging station i, let the number of its charging devices, the number of new energy units, and the number of energy storage systems be n char,i , n RES,i and n BESS,i , respectively. The corresponding reliability indexes include EENS, MTTR, LOLP, SAIDI, and SAIFI. Then, a spline interpolation function S z is established:

[0068] S z (n char,i , n RES,i , n BESS,i )(21)

[0069] And the known data points are

[0070]

[0071] represents that under the kth configuration , the observed index value is z k ; assuming that m segmentation points are used in each dimension, then S z is expressed as:

[0072]

[0073] Among them, B j , B k and B l are the basis spline functions at the nodes, describing the piecewise polynomials in each dimension; c jkl is the coefficient to be fitted, representing the weight of each basis spline combination;

[0074] The piecewise connection points of the spline interpolation function should satisfy the following smoothness conditions:

[0075]

[0076] Among them, x - and x + represent the left and right sides of the piecewise point respectively;

[0077] The coefficients c of the piecewise polynomial are solved by minimizing the following squared error objective function jkl :

[0078]

[0079] The solution is obtained by linear algebra or numerical methods, and the result is the weight c of each basis spline jkl ;

[0080] Assume that the values of two dimensions are fixed, and spline interpolation is performed on the third dimension:

[0081]

[0082] In the above formula, they are the relational expressions for fitting one variable while fixing the other two variables respectively. Based on the above calculations, the final fitted form of the spline interpolation function is:

[0083] z = S z (n char,i , n RES,i , n BESS,i ) (27)

[0084] Among them, S z is a combination of piecewise polynomials.

[0085] Preferably, the reliability improvement model in step S5 is expressed as:

[0086]

[0087] Satisfying:

[0088]

[0089] Among them, and Denote the costs of the charging equipment, energy storage system, and new energy unit of charging station i; decomposed as:

[0090]

[0091] R i Denote the weight of the i-th charging station; L r Denote the set of charging stations; and Denote the number of groups of the charging equipment, energy storage system, and new energy unit; c char c BESS and c char Denote the unit prices of the charging equipment, energy storage system, and new energy unit; δ k,p,q is a piecewise activation variable used to activate the corresponding fitting function in different intervals;

[0092] a k,p,q +b k,p,q n char,i +c k,p,q n RES,i +d k,p,q n BESS,i is the expression of the spline function in different intervals; z target,i is the requirement for assuming all reliability indicators; k, p, q represent different segments; m x m y and m z Denote the set of different segments; (x k , x k+1 ), (y p , y p+1 ), and (z q , z q+1 ) represent the segment intervals corresponding to different components respectively; M is the big M method to convert the segment logic into linear constraints; and Denote the minimum installed capacity of each charging station i; and Denote the maximum installed capacity of each charging station i.

[0093] Therefore, when the above method for improving the reliability of the offshore charging station considering time-varying faults and new energy randomness is adopted in the present invention, the following beneficial effects are obtained:

[0094] (1) Establish the failure rate models (exponential distribution, Weibull distribution, normal distribution) of the charging equipment, energy storage system, and new energy unit, and generate the time series data of the equipment state by combining Monte Carlo sampling, realizing the coupling simulation of endogenous faults and exogenous environmental disturbances, and improving the comprehensiveness and accuracy of the reliability assessment;

[0095] (2) A reliability-redundancy surface fitting method based on spline interpolation is proposed to convert discrete Monte Carlo simulation data into a continuous and differentiable analytical function, quantify the non-linear relationship between redundant configurations (number of charging devices, energy storage capacity, fan redundancy) and reliability indicators (EENS, LOLP, SAIDI), break through the limitations of traditional empirical capacity expansion, and provide a computable mathematical basis for redundancy optimization;

[0096] (3) A multi-scenario dynamic Monte Carlo sampling framework is designed. By simulating the combined effects of load fluctuations, intermittency of new energy output, and energy storage charging and discharging strategies, an operation scenario library covering normal, extreme weather, and high load conditions is generated to ensure that the reliability assessment covers all operating condition risks;

[0097] (4) An optimization model with the goal of minimizing equipment cost and reliability indicators as constraints is constructed. Utilizing the differentiable characteristics of the spline surface, rapid gradient optimization of redundant configurations is achieved, and the construction cost of charging stations is reduced on the premise of meeting the reliability threshold, solving the problem of excessively high computational complexity in the collaborative optimization of traditional Monte Carlo simulation and planning models.

[0098] The technical solution of the present invention will be further described in detail below through the accompanying drawings and embodiments. Description of the Drawings

[0099] Figure 1 It is a flowchart of the method for improving the reliability of an offshore charging station considering time-varying faults and new energy randomness according to the present invention. Detailed Embodiments

[0100] The following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed present invention, but merely represents selected embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts fall within the scope of protection of the present invention.

[0101] Please refer to Figure 1 , a method for improving the reliability of an offshore charging station considering time-varying faults and new energy randomness, includes the following steps:

[0102] S1. Construction of Evaluation Index

[0103] To evaluate the power supply reliability of an electric ship charging station, appropriate evaluation indicators need to be established. For a charging station, the power supply reliability of the charging station is measured by the Expected Energy Not Supplied (EENS) and the Mean Time to Repair (MTTR). EENS refers to the expected value of the electric energy that fails to be supplied to the load due to system power outages or equipment failures. It reflects the degree to which the system fails to meet the charging requirements of electric ships under specific operating conditions and is usually expressed in electric energy units (such as kWh or MWh). The average repair time MTTR reflects the average time required for the equipment to return to an available state after a failure, reflecting the system's fault response efficiency and operation and maintenance capabilities. For an electric ship charging station, a lower MTTR value means that when the equipment has problems, the charging station can quickly restore its power supply function, thus reducing the impact on the charging task. These two indicators are respectively expressed as:

[0104]

[0105]

[0106] Among them, is the weight of each charging station on the route, determined according to the installed capacity of each charging station; represents the load power of charging station i at time t; represents the charging power of the i-th charging station at time t; n is the total number of charging stations; Δt is the time interval; represents the repair time of the f-th failure of the i-th charging station; represents the total number of failures of the i-th charging station; F i the failure set of the i-th charging station;

[0107] For electric ships, insufficient reliability will lead to an increase in the queuing time of electric ships, and frequent queuing will affect shipping efficiency. This problem can be evaluated using the Loss of Load Probability (LOLP) indicator. LOLP refers to the probability that the system load demand cannot be met, usually expressed as a probability value over a certain time period (such as one year). It is a direct reflection of the balance between the system's power supply capacity and load demand. It can be expressed as:

[0108]

[0109] Among them, is an indicator function. If the load demand of the i-th charging station is greater than the available power at time t, it takes the value of 1, otherwise it is 0.

[0110] Longer power outage duration will lead to an increase in the queuing time of ships. By analogy with the SAIDI (System Average Interruption Duration Index) index in the distribution network, the average duration of insufficient power supply for electric ships can be calculated to reflect the additional queuing time of electric ships. This index can be expressed as:

[0111]

[0112] where U i,t represents the number of users affected by the power outage at the i-th charging station at time t;

[0113] Another index in the distribution network, SAIFI (System Average Interruption Frequency Index), refers to the occurrence frequency of power outage events. It can also be used to count the average number of power outages experienced by each ship per year, reflecting the frequency of power outage events. For electric ships, frequent power outages may lead to multiple interruptions of the charging plan, increasing the uncertainty of ship operation. This index is expressed as:

[0114]

[0115] where α i,t is an indicator function. If the i-th charging station is operating normally at time t, its value is 1; otherwise, it is 0; is the total number of users of the i-th charging station.

[0116] S2. Reliability Evaluation Model for Electric Ship Charging Stations

[0117] For offshore charging stations, the failure rate model of charging equipment is considered first. Charging equipment includes charging interfaces, insulation materials, and contactors, and mainly suffers from physical wear, current shock, or damage caused by occasional external factors. In actual operation, the failure rate of the charging head is often assumed to be constant, without significant "early failure" or "aging effect", so it is suitable to be modeled using the exponential distribution. Therefore, its failure distribution f t cf can be expressed as:

[0118] f t cf = λe -λt (6)

[0119] where λ is the failure rate constant;

[0120] The failures of energy storage systems (such as lithium batteries) are usually related to aging, and their failure rates change over time. The failure rate is relatively low in the initial stage (running-in period), stable in the middle stage, and significantly increases in the later stage (decline period). The Weibull distribution can approximately describe this characteristic through the shape parameter β and the scale parameter η. Therefore, the failure distribution of the energy storage system is expressed as:

[0121]

[0122] Most of the mechanical components of wind turbines (such as blades, gearboxes, and bearings) have a clear design life. Failures mostly occur in the middle and later stages of the life, and the relative distribution is relatively uniform. The normal distribution represents the average failure life of the wind turbine through the mean μ, and the standard deviation σ describes the discreteness of the failure time. For equipment like wind turbines with relatively clear life characteristics, the normal distribution can better depict the central tendency of its failure. Therefore, the failure distribution of the wind turbine can be expressed as:

[0123]

[0124] The above derivation obtains the failure distribution of each component. Next, the temporal distribution of component failures is inferred. For all components, O t is used to represent its failure state at time t. If it fails at time t, then O t = 0; otherwise, it is 1. Since the failure state is related to time, the failure situation at the next moment is inferred based on the previous moment, which is expressed as:

[0125] P(O t+1 = 0|O t = 1) = f t (9)

[0126] Among them, P(O t+1 = 0|O t = 1) represents the probability that the component is still faulty at time t + 1 when it is operating normally at time t. If it fails, its repair probability is p re , which is expressed as:

[0127] P(O t+1 = 1|O t = 0) = p re (10)

[0128] In fact, the repair cannot be completed immediately and there is a certain time interval, assumed to be T re , then the formulas (9) and (10) are corrected to obtain:

[0129]

[0130] S3. System operation model based on Monte Carlo sampling

[0131] First, perform Monte Carlo sampling on the device failure state to generate a failure state time series. First, define the system component set C = {c1, c2,... c n}, the total number of simulation times M, and the time series T = {t1, t2,... t k}, where k is the number of time steps;

[0132] For each component c i , randomly generate a set of samples based on its failure distribution; for the exponential distribution, let the failure rate constant be λ, and generate the sampling time through the inverse transformation method:

[0133]

[0134] For the Weibull distribution, let the scale parameter and shape parameter be η and β respectively, and perform inverse sampling according to its cumulative distribution function:

[0135]

[0136] For the normal distribution, generate normal samples randomly:

[0137]

[0138] After obtaining the sampling time, gradually update the component state within the time series T; if the current time then the component c i is marked as the failure state until the repair is completed;

[0139] After obtaining the Monte Carlo sampling data, substitute the component failure state into the system operation simulation model. For each charging station i, the output of its new energy unit is corrected to:

[0140]

[0141] where j represents the number of the new energy unit in the charging station i; N j represents the total number of new energy units; is the failure state of each unit; represents the output of each unit; and the remaining power of the new energy unit is:

[0142]

[0143] where L i,t represents the charging demand of the charging station at this moment; assume that the discharge power of the energy storage system is then:

[0144]

[0145] where, For the fault state of the energy storage system; after obtaining the discharge power, calculate its energy S i,j,t Variation:

[0146]

[0147] Similarly, obtain the power situation and energy variation of the energy storage system during charging:

[0148]

[0149]

[0150] Among them, η dis represents the discharge efficiency; represents the maximum capacity of the energy storage system; η ch represents the charging efficiency;

[0151] According to the above derivation, obtain the total output of the new energy unit and the energy storage system, and calculate its power gap.

[0152] S4. Method for improving the reliability of electric ship charging stations

[0153] There is a trade-off between reliability and redundancy. In the case of low redundancy, the reliability of the system is significantly insufficient, resulting in frequent interruptions during the charging process, which in turn affects the normal operation of electric ships; after the redundancy exceeds a certain level, its marginal improvement effect on reliability will gradually weaken, while the cost continues to increase. Therefore, it is necessary to find a redundant configuration of the charging station that can both meet the reliability requirements and control the cost.

[0154] Select spline interpolation as the fitting method for this study. The fitting process is as follows:

[0155] For charging station i, let the number of its charging devices, the number of new energy units, and the number of energy storage systems be n char,i , n RES,i and n BESS,i , and the corresponding reliability indexes include EENS, MTTR, LOLP, SAIDI, and SAIFI. Then establish a spline interpolation function S z :

[0156] S z (n char,i , n RES,i , n BESS,i )(21)

[0157] And the known data points are

[0158]

[0159] Denote the observed metric value as z under the k-th configuration ; The research needs to find a function S k that is continuously differentiable in three-dimensional space, can accurately fit the known data points, and has good smoothness. To describe S in three-dimensional space z , it is expressed as a combination of three-dimensional piecewise polynomials. Assume m piecewise points are used in each dimension, then S z is expressed as: z where, B

[0160]

[0161] j k l are the basis spline functions at the nodes, describing the piecewise polynomials in each dimension; c jkl are the coefficients to be fitted, representing the weights of each basis spline combination;

[0162] To ensure the continuity and differentiability of the function, the following smoothness conditions should be satisfied at the piecewise joints of the spline interpolation function:

[0163]

[0164] where, x - and x + represent the left and right sides of the piecewise points respectively; The above equations ensure that the function has continuous values, first-order derivatives, and second-order derivatives at the piecewise joints.

[0165] The core of spline interpolation is to determine the coefficients c jkl of the piecewise polynomials. To ensure that the function fits the known data points as much as possible while avoiding overfitting, it is usually solved by minimizing the following squared error objective function:

[0166]

[0167] This optimization problem is solved by linear algebra or numerical methods, and the result is the weight c jkl of each basis spline;

[0168] Since directly dealing with three-dimensional spline fitting is too complex, the fitting can be carried out step by step for a single dimension. Assume that the values of two dimensions are fixed, and spline interpolation is performed on the third dimension:

[0169]

[0170] In the above equation, they are the relational expressions for fitting one variable while fixing the other two variables. Based on the above calculations, the final fitted form of the spline interpolation function is:

[0171] z = S​​​z (n char,i ,n RES,i ,n BESS,i ) (27)

[0172] Among them, S z is a combination of piecewise polynomials.

[0173] S5. The objective function for reliability improvement based on the charging reliability - redundancy surface is the construction cost of the charging station under different redundancy configurations, expressed as:

[0174]

[0175] Among them, R i represents the weight of the i-th charging station; L r represents the set of charging stations; and represent the costs of the charging equipment, energy storage system, and new energy unit of charging station i; the costs of various equipment can be disassembled as:

[0176]

[0177] Among them, and represent the number of sets of charging equipment, energy storage system, and new energy unit; since the capacity of each charging equipment, energy storage system, and new energy unit used in different charging stations is equal, the difference in the total installed capacity lies in the number of sets. Therefore, c char , c BESS and c char can be used to represent the unit prices of charging equipment, energy storage system, and new energy unit;

[0178] The constraints of the reliability improvement model include the upper and lower bounds of reliability constraints and the upper and lower bounds of installed capacity constraints. The upper and lower bounds of reliability constraints specify the expected reliability requirements, and the upper and lower bounds of installed capacity constraints specify the upper and lower limits of the installed capacity that can be invested. The reliability fitting function is expressed as:

[0179]

[0180] Among them, δ k,p,q is a piecewise activation variable used to activate the corresponding fitting function in different intervals; a k,p,q +b k,p,q n char,i +c k,p,q n RES,i +d k,p,q n BESS,i is the expression of the spline function in different intervals; and the activation variable needs to satisfy:

[0181]

[0182] The above equation indicates that only one segment is activated at any given time. Assume that all reliability index requirements are z target,i , then the constraint can be expressed as:

[0183]

[0184] For each segment (k, p, q), it is necessary to ensure that the activation variable δ k,p,q corresponds to the current interval, that is:

[0185]

[0186] where (x k , x k+1 ), (y p , y p+1 ) and (z q , z q+1 ) represent the segment intervals corresponding to different components respectively. In order to convert the segment logic into linear constraints, it is modified to:

[0187]

[0188] The above equation converts the segment logic into linear constraints through the big M method;

[0189] The upper and lower limits of the installed capacity constraint can be expressed as:

[0190]

[0191] where, and represent the minimum installed capacity of each charging station i. In this study, the installed capacity optimized without considering failures is selected as the minimum installed capacity; and represent the maximum installed capacity of each charging station i, which is set according to the affordable cost.

[0192] To sum up, the reliability improvement model can be expressed as:

[0193]

[0194] Subject to:

[0195]

[0196] Therefore, when considering the above time-varying faults and the randomness of new energy, the reliability improvement method of the offshore charging station of the present invention establishes an explicit mathematical relationship between the redundant configuration of charging equipment, energy storage systems, and new energy units and reliability indicators. By using spline interpolation to fit the discrete reliability data generated by Monte Carlo simulation into a continuously differentiable surface, an optimization model with the goal of minimizing equipment costs and reliability indicators as constraints is constructed. Combining Monte Carlo dynamic sampling to generate multi-scenario operation data, rapid gradient optimization of redundant configuration is achieved, reducing construction costs while ensuring reliability.

[0197] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that they can still modify or equivalently replace the technical solutions of the present invention, and these modifications or equivalent replacements do not make the modified technical solutions deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for improving the reliability of offshore charging stations considering time-varying faults and randomness of new energy sources, characterized in that: The following steps are involved: S1. Construct an evaluation index for evaluating the reliability of power supply for electric ship charging stations; S2. Construct a reliability assessment model for electric boat charging stations, including component reliability and system reliability; S3. Generate a fault status time series based on Monte Carlo sampling, and gradually update the component status within the time series until the repair is completed, thereby obtaining the reliability assessment result of the offshore charging station; S4. Perform redundant configuration of charging stations based on reliability requirements and cost control, and fit the evaluation index of S1 with the reliability evaluation result of S3; S5. Based on reliability requirements, establish a reliability improvement model.

2. The method for improving the reliability of offshore charging stations considering time-varying faults and randomness of new energy sources according to claim 1 is characterized in that: The evaluation indicators of step S1 include: For charging stations, the power supply reliability of charging stations is measured by the expected amount of unavailable electricity EENS and the mean repair time MTTR, which are expressed as: in, The weight of each charging station on the route is determined according to the installed capacity of each charging station; represents the load power of charging station i at time t; represents the charging power of the i-th charging station at time t; n is the total number of charging stations; Δt is the time interval; represents the repair time of the fth fault of the i-th charging station; represents the total number of failures of the i-th charging station; F i The fault set of the i-th charging station; For electric ships, the reliability is measured by the load failure probability LOLP, the system average power outage duration index SAIDI and the system average power outage frequency index SAIFI, which are expressed as: in, is an indicator function. If the load demand of the i-th charging station is greater than the available power at time t, the value is 1, otherwise it is 0; U i,t represents the number of users affected by the power outage at the i-th charging station at time t; α i,t is the indicator function, which takes the value of 1 if the i-th charging station operates normally at time t, otherwise it takes the value of 0; is the total number of users at the i-th charging station.

3. The method for improving the reliability of offshore charging stations considering time-varying faults and randomness of new energy sources according to claim 2 is characterized in that: The component reliability assessment model of step S2 includes: The failure rate model of charging equipment is modeled using exponential distribution, and the failure distribution is expressed as: f t cf =λe -λt (6) Where, λ is the failure rate constant; The energy storage system is modeled using Weibull distribution, and its fault distribution is expressed as: Among them, β is the shape parameter; η is the scale parameter; The offshore wind turbine is modeled using normal distribution, and the wind turbine failure distribution is expressed as; Among them, σ is the standard deviation, which is used to describe the discrete type of failure time; μ is the mean, which represents the average failure life of the fan.

4. The method for improving the reliability of offshore charging stations considering time-varying faults and randomness of new energy sources according to claim 3 is characterized in that: The system reliability evaluation model in step S2 adopts t Indicates its fault state at time t. If it fails at time t, then O t =0, otherwise 1. Since the fault state is related to time, the fault condition at the next moment is inferred based on the previous moment, expressed as: NIGHT t+1 =0|O t =1)=f t (9) Among them, P(O t+1 =0|O t =1) indicates the probability that a component is still faulty at time t+1 if it is operating normally at time t; if it fails, the probability of repair is p re , expressed as: AFTER t+1 =1|O t =0)=p re (10) Assume that the repair time interval is T re , then we can modify formulas (9) and (10) to obtain:

5. The method for improving the reliability of offshore charging stations considering time-varying faults and randomness of new energy sources according to claim 4 is characterized in that: Step S3 specifically includes: S31, performing Monte Carlo sampling on the equipment fault state to generate a fault state time series; S32. Get sampling time After that, the component status is gradually updated in the time series; if the current time Then component c i It is marked as a faulty state until repair is completed; S33. After obtaining the Monte Carlo sampling data, the component failure status is substituted into the system operation simulation model to obtain the total output of the new energy unit and the energy storage system, and calculate its power gap.

6. The method for improving the reliability of offshore charging stations considering time-varying faults and randomness of new energy sources according to claim 5 is characterized in that: Step S31 is specifically as follows: First, define the system component set C = {c1, c2, ...c n }, total number of simulations M, time series T = {t1, t2, ...t k }, where k is the number of time steps; For each component c i , randomly generate a set of samples based on its fault distribution; for exponential distribution, generate sampling time by inverse transformation method: For the Weibull distribution, backsample according to its cumulative distribution function: For a normal distribution, generate normal samples randomly:

7. The method for improving the reliability of offshore charging stations considering time-varying faults and randomness of new energy sources according to claim 6 is characterized in that: Step S33 is specifically as follows: For each charging station i, the output of its new energy unit is corrected as follows: Where j represents the number of the new energy unit in charging station i; N j Indicates the total number of new energy units; The fault status of each unit; Indicates the output of each unit; and the remaining power of the new energy unit is: Among them, L i,t represents the charging demand of the charging station at that moment; assuming that the discharge power of the energy storage system is but: in, is the fault state of the energy storage system; after obtaining the discharge power, its energy S is calculated i,j,t Changes: Similarly, the power and energy changes of the energy storage system charging are obtained: Among them, η dis Indicates discharge efficiency; Represents the maximum capacity of the energy storage system; η ch Indicates charging efficiency; Based on the above derivation, the total output of the new energy units and energy storage system is obtained, and its power gap is calculated.

8. The method for improving the reliability of offshore charging stations considering time-varying faults and randomness of new energy sources according to claim 7 is characterized in that: In step S4, spline interpolation is used as the fitting method, and the fitting process is as follows: Assume that for charging station i, the number of charging equipment, the number of new energy units and the number of energy storage systems are n char,i 、n RES,i and n BESS,i , the corresponding reliability indicators include EENS, MTTR, LOLP, SAIDI and SAIFI, then the spline interpolation function S describing the reliability indicators is established z : S z (n char,i ,n RES,i ,n BESS,i ) (21) The known data points are Indicates that in the kth configuration Under this condition, the observed indicator value is z k ; Assuming that m segmentation points are used in each dimension, then S z It is expressed as: Among them, B j , B k and B l is the basis spline function at the nodes, describing the piecewise polynomial in each dimension; c jkl is the coefficient to be fitted, which represents the weight of each basis spline combination; The segmented connections of the spline interpolation function should satisfy the following smoothness conditions: Among them, x - and x + Respectively represent the left and right sides of the segmentation point; The coefficients c of the piecewise polynomial are solved by minimizing the following squared error objective function jkl : Solved by linear algebra or numerical methods, the result is the weight c of each basis spline jkl ; Assume that the values ​​of two dimensions are fixed and the spline interpolation is performed on the third dimension: In the above formula, the other two variables are fixed and one of the variables is fitted. Based on the above calculation, the final fitted spline interpolation function is in the form of: z=S z (n char,i ,n RES,i ,n BESS,i ) (27) Among them, S z is a combination of piecewise polynomials.

9. The method for improving the reliability of offshore charging stations considering time-varying faults and randomness of new energy sources according to claim 8 is characterized in that: The reliability improvement model of step S5 is expressed as: Satisfied with: in, and Represents the cost of charging equipment, energy storage system and new energy unit of charging station i; it is broken down into: R i represents the weight of the i-th charging station; Lr Represents a collection of charging stations; and Indicates the number of charging equipment, energy storage systems and new energy units; c char 、c BESS and c char Indicates the unit price of charging equipment, energy storage system and new energy unit; δ k,p,q is a piecewise activation variable, which is used to activate the corresponding fitting function in different intervals; a k,p,q +b k,p,q n char,i +c k,p, q n RES,i +d k,p,q n BESS,i is the expression of spline function in different intervals; z target,i is the assumption that all reliability indicators are required; k, p, q represent different segments; m x 、m y and m z Represents a collection of different segments; (x k ,x k+1 ), (y p ,y p+1 ) and (z q ,z q+1 ) represent the segmented intervals corresponding to different components; M is the big M method that converts the segmented logic into linear constraints; and represents the minimum installed capacity of each charging station i; and represents the maximum installed capacity of each charging station i.

Images