Risk assessment method for failure mode of photovoltaic power generation system

Abstract

The invention discloses a risk assessment method for a failure mode of a photovoltaic power generation system, and belongs to the technical field of renewable energy sources. The method comprises the following steps of (1) determining a language Z number decision matrix, (2) converting each piece of language Z number evaluation information in the decision matrix into a rough-Z number, (3) calculating each evaluation index weight, (4) constructing a rough-Z number decision matrix, and (5) selecting a fault with the maximum risk based on an improved failure mode and influence analysis method. According to the method, the complexity of potential fault identification and risk assessment of the photovoltaic power generation system can be effectively solved, and the problem that subjectivity, uncertainty and reliability of evaluation information are difficult to process in an existing failure mode and influence analysis method is solved; the accuracy and the decision-making efficiency of potential fault identification of a decision maker in the operation process of the photovoltaic power generation system are remarkably improved, and the safety and the reliability of system operation are guaranteed.

Description

Technical Field

[0001] The invention belongs to the technical field of renewable energy, and in particular relates to a risk assessment method for failure modes of a photovoltaic power generation system. Background Art

[0002] With the widespread application of photovoltaic power generation technology, photovoltaic power generation system components and balance of system (BOS) and other components face multiple potential faults during operation, such as component packaging material delamination, cell cracking, hot spots, cable disconnection, inverter and transformer failure, etc., threatening the stability and safety of system operation. There are large differences in the severity, incidence, difficulty of detection, maintenance costs and technical difficulty of different potential faults. When identifying potential faults, operating companies often face the problem of how to balance reducing the risk of photovoltaic power generation systems and improving the quality of operation and maintenance.

[0003] Existing potential fault identification methods mostly use failure mode and effect analysis algorithms based on multi-attribute decision-making. However, due to the limited domain knowledge and experience of decision-makers, existing methods have problems such as difficulty in accurately measuring subjective information and complex calculation processes when dealing with complex decision-making information of photovoltaic power generation systems. At the same time, existing failure mode and effect analysis methods only evaluate the severity, occurrence rate and difficulty of detection of potential faults, and are relatively limited in considering the difficulty of fault repair. In addition, the application of existing failure mode and effect analysis methods in potential fault identification and risk assessment of photovoltaic power generation systems is relatively limited, and it is difficult to meet the stability and reliability requirements of power generation systems. Summary of the invention

[0004] The present invention aims to solve the above problems in the existing failure mode and effect analysis methods, and provides a risk assessment method for failure modes of photovoltaic power generation systems, which is used for potential fault identification and risk assessment during the operation of photovoltaic power generation systems. The method can comprehensively consider the differences in severity, occurrence, difficulty of detection, maintenance costs and technical difficulty of potential faults, characterize the subjectivity and reliability of evaluation information through rough numbers and language Z numbers, and propose and use the improved MULTIMOORA method to finally select the potential faults with the greatest risk.

[0005] The technical solution adopted by the present invention is a risk assessment method for failure modes of photovoltaic power generation systems, the method comprising the following steps:

[0006] Step 1: Determine the Language Z Decision Matrix

[0007] In this step, we first need to determine the potential faults of the photovoltaic power generation system and their evaluation indicators, collect the evaluation information and express this information in the form of a language Z number; the language Z number is usually composed of two parts, A and B, that is, Z = (A, B), where:

[0008] represents the restriction on the evaluation value, is the triangular membership function of A;

[0009] represents the reliability measure of evaluating A, is the triangular membership function of B;

[0010] The importance evaluation value of each indicator can be expressed by the language Z number as shown below:

[0011]

[0012] in, represents the k-th language Z-number evaluation value under the j indicator, k belongs to the evaluation times set, which contains t evaluations in total, and j belongs to the indicator set;

[0013] The attribute value of each potential fault under each indicator can also be expressed by the language Z number and form a decision matrix as shown below:

[0014] D=(Z ij ) m×n

[0015] Among them, Z ij =(A ij , B ij ) represents the language Z-number evaluation of the i-th potential fault under the j-th index, i belongs to the potential fault set, there are m potential faults in total, j belongs to the index set, there are n indexes in total;

[0016] Step 2: Convert the Z-score evaluation information of each language in the decision matrix into a rough Z-score

[0017] Since the language Z number cannot fully represent the information uncertainty in the evaluation process, the value of the language Z number needs to be converted into a rough-Z number through the following formula. The specific operation method is:

[0018] (1) The linguistic Z number Z = (A, B) is converted into a triangular fuzzy number Z' = (a, b, c) by the following three formulas:

[0019]

[0020] (2) The triangular fuzzy number (a, b, c) is converted into a rough Z number RZ(Z) by the following steps: - , a + ],[b - , b + ],[c - , c + ])

[0021] First, determine the lower and upper bases of a, b, and c:

[0022] Apr (a) = ∪{R∈U / Ξ a (R)≤a}

[0023] Apr (b) = ∪{R∈U / Ξ b (R)≤b}

[0024] Apr (c) = ∪{R∈U / Ξ c (R)≤c}

[0025]

[0026] Then calculate the lower and upper limits of the rough numbers of a, b, and c respectively:

[0027]

[0028] Where U is a set of natural numbers, and R is any element in U. a ,Ξ b ,Ξ c are the sets of a, b, and c respectively. and Respectively represent the lower limit set Apr (a) Apr (b) and Apr (c) The number of elements in the same way, and Respectively represent the upper limit base and The number of elements in ;

[0029] Step 3: Calculate the weight of each evaluation indicator

[0030] In potential fault analysis, it is crucial to determine the indicator weights. Therefore, in order to determine the degree of influence of each indicator on potential faults, in this step, the following two formulas are used to derive the standard weights of each evaluation indicator;

[0031]

[0032] in, represents the standard weight of the jth standard;

[0033] Step 4: Construct the rough-Z number decision matrix

[0034] Since the rough-Z number can comprehensively represent the reliability and uncertainty of the evaluation information, the rough-Z number decision matrix is ​​constructed according to the rough-Z number evaluation calculated in the previous article;

[0035] The rough Z number decision matrix can be expressed as: RZ = (RZ ij ) m×n

[0036] in,

[0037] represents the roughness-z number evaluation of the ith potential fault under the jth index;

[0038] Step 5: Select the fault with the greatest risk based on the improved FMEA method

[0039] In this step, the improved MULTIMOORA method is used to calculate the comprehensive risk value of each potential fault, and then the fault with the greatest risk is selected;

[0040] (1) Obtain the risk value of each potential failure based on the ratio system analysis method:

[0041]

[0042] Among them, y i represents the risk value of the ith potential failure, j = 1, 2, ..., h is a benefit index, j = h+1, h+2, ..., n is a cost index; in this method, y i The larger the value, the greater the potential failure risk.

[0043] (2) Obtain the risk value of each potential fault based on the reference point method:

[0044]

[0045] Among them, r j represents the reference point of the jth indicator, d ij Indicates the deviation between the evaluation value and the reference point, z i is the risk value of the ith potential failure; z i The smaller it is, the greater the potential failure risk is;

[0046] (3) Obtain the risk value of each potential fault based on the full multiplication form:

[0047]

[0048] Among them, μ i represents the risk value of the ith potential failure, μ i The higher the value, the greater the potential failure risk;

[0049] (4) Calculate the final risk value of each potential failure:

[0050]

[0051] Among them, R(i) represents the final risk value of the ith potential failure, r(y i ), r(z i ), r(μ i ) represent the ranking of the i-th potential fault under the three methods respectively; the larger R(i) is, the higher the ranking is and the greater the risk is.

[0052] Beneficial effects of the present invention

[0053] The risk assessment method for failure modes of photovoltaic power generation systems proposed in the present invention can effectively solve the complexity of potential fault identification and risk assessment of photovoltaic power generation systems, overcome the problem of difficulty in handling the subjectivity, uncertainty and reliability of evaluation information in existing failure mode and impact analysis methods, and significantly improve the accuracy and decision-making efficiency of decision makers in identifying potential faults during the operation of photovoltaic power generation systems, thereby ensuring the safety and reliability of system operation. The present invention can also be applied to the problem of potential fault identification and risk assessment of photovoltaic power generation enterprise system operation in an uncertain information environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 It is a risk assessment method for failure modes of photovoltaic power generation systems. DETAILED DESCRIPTION

[0055] like Figure 1 As shown, the risk assessment method for failure modes of a photovoltaic power generation system described in this embodiment includes the following steps:

[0056] Step 1: Determine the Language Z Decision Matrix

[0057] In this step, we first need to determine the potential faults of the photovoltaic power generation system and their evaluation indicators, collect the evaluation information and express this information in the form of a language Z number; the language Z number is usually composed of two parts, A and B, that is, Z = (A, B), where:

[0058] represents the restriction on the evaluation value, is the triangular membership function of A;

[0059] represents the reliability measure of evaluating A, is the triangular membership function of B;

[0060] The importance evaluation value of each indicator can be expressed by the language Z number as shown below:

[0061]

[0062] in, represents the k-th language Z-number evaluation value under the j indicator, k belongs to the evaluation times set, which contains t evaluations in total, and j belongs to the indicator set;

[0063] The attribute value of each potential fault under each indicator can also be expressed by the language Z number and form a decision matrix as shown below:

[0064] D=(Z ij ) m×n

[0065] Among them, Z ij =(A ij , B ij ) represents the language Z-number evaluation of the i-th potential fault under the j-th index, i belongs to the potential fault set, there are m potential faults in total, j belongs to the index set, there are n indexes in total;

[0066] Step 2: Convert the Z-score evaluation information of each language in the decision matrix into a rough Z-score

[0067] Since the language Z number cannot fully represent the information uncertainty in the evaluation process, the value of the language Z number needs to be converted into a rough-Z number through the following formula. The specific operation method is:

[0068] (1) The linguistic Z number Z = (A, B) is converted into a triangular fuzzy number Z' = (a, b, c) by the following three formulas:

[0069]

[0070] (2) The triangular fuzzy number (a, b, c) is converted into a rough Z number RZ(Z) by the following steps: - , a + ],[b - , b + ],[c - , c + ])

[0071] First, determine the lower and upper bases of a, b, and c:

[0072] Apr (a) = ∪{R∈U / Ξ a (R)≤a}

[0073] Apr (b) = ∪{R∈U / Ξ b (R)≤b}

[0074] Apr (c) = ∪{R∈U / Ξ c (R)≤c}

[0075]

[0076] Then calculate the lower and upper limits of the rough numbers of a, b, and c respectively:

[0077]

[0078]

[0079] Where U is a set of natural numbers, and R is any element in U; a ,Ξ b ,Ξ c are the sets of a, b, and c respectively. and Respectively represent the lower limit set Apr (a) Apr (b) and Apr (c) The number of elements in the same way, and Respectively represent the upper limit base and The number of elements in ;

[0080] Step 3: Calculate the weight of each evaluation indicator

[0081] In potential fault analysis, it is crucial to determine the indicator weights. Therefore, in order to determine the degree of influence of each indicator on potential faults, in this step, the following two formulas are used to derive the standard weights of each evaluation indicator;

[0082]

[0083] in, represents the standard weight of the jth standard;

[0084] Step 4: Construct the rough-Z number decision matrix

[0085] Since the rough-Z number can comprehensively represent the reliability and uncertainty of the evaluation information, the rough-Z number decision matrix is ​​constructed according to the rough-Z number evaluation calculated in the previous article;

[0086] The rough Z number decision matrix can be expressed as: RZ = (RZ ij ) m×n

[0087] in,

[0088] represents the roughness-z number evaluation of the ith potential fault under the jth index;

[0089] Step 5: Select the fault with the greatest risk based on the improved FMEA method

[0090] In this step, the improved MULTIMOORA method is used to calculate the comprehensive risk value of each potential fault, and then the fault with the greatest risk is selected;

[0091] (1) Obtain the risk value of each potential failure based on the ratio system analysis method:

[0092]

[0093] Among them, y i represents the risk value of the ith potential failure, j = 1, 2, ..., h is a benefit index, j = h+1, h+2, ..., n is a cost index; in this method, y i The larger the value, the greater the potential failure risk.

[0094] (2) Obtain the risk value of each potential fault based on the reference point method:

[0095]

[0096] Among them, r j represents the reference point of the jth indicator, d ij Indicates the deviation between the evaluation value and the reference point, z i is the risk value of the ith potential failure; z i The smaller it is, the greater the potential failure risk is;

[0097] (3) Obtain the risk value of each potential fault based on the full multiplication form:

[0098]

[0099] Among them, μ i represents the risk value of the ith potential failure, μ i The higher the value, the greater the potential failure risk;

[0100] (4) Calculate the final risk value of each potential failure:

[0101]

[0102] Among them, R(i) represents the final risk value of the ith potential failure, r(y i ), r(z i ), r(μ i ) represent the ranking of the i-th potential fault under the three methods respectively; the larger R(i) is, the higher the ranking is and the greater the risk is.

Claims

1. A risk assessment method for failure modes of photovoltaic power generation systems, characterized in that: The method comprises the following steps: Step 1: Determine the Language Z Decision Matrix In this step, we first need to determine the potential faults of the photovoltaic power generation system and their evaluation indicators, collect the evaluation information and express this information in the form of a language Z number; the language Z number is usually composed of two parts, A and B, that is, Z = (A, B), where: represents the restriction on the evaluation value, is the triangular membership function of A; represents the reliability measure of evaluating A, is the triangular membership function of B; The importance evaluation value of each indicator can be expressed by the language Z number as shown below: in, represents the k-th language Z-number evaluation value under the j indicator, k belongs to the evaluation times set, which contains t evaluations in total, and j belongs to the indicator set; The attribute value of each potential fault under each indicator can also be expressed by the language Z number and form a decision matrix as shown below: D=(Z ij ) m×n Among them, Z ij =(A ij , B ij ) represents the language Z-number evaluation of the i-th potential fault under the j-th index, i belongs to the potential fault set, there are m potential faults in total, j belongs to the index set, there are n indexes in total; Step 2: Convert the Z-score evaluation information of each language in the decision matrix into a rough Z-score Since the language Z number cannot fully represent the information uncertainty in the evaluation process, the value of the language Z number needs to be converted into a rough-Z number through the following formula. The specific operation method is: (1) The linguistic Z number Z = (A, B) is converted into a triangular fuzzy number Z' = (a, b, c) by the following three formulas: (2) The triangular fuzzy number (a, b, c) is converted into a rough Z number by the following steps: RZ(Z)=([a - a + ],[b - ,b + ],[c - ,c + ]) First, determine the lower and upper bases of a, b, and c: April (a)=∪{R∈U / Ξ a (R)≤a} April (b)=∪{R∈U / Ξ b (R)≤b} April (c)=∪{R∈U / Ξ c (R)≤c} Then calculate the lower and upper limits of the rough numbers of a, b, and c respectively: Where U is a set of natural numbers, and R is any element in U; a ,Ξ b ,Ξ c are the sets of a, b, and c respectively. and Respectively represent the lower limit set April (a) April (b) and April (c) The number of elements in the same way, and Respectively represent the upper limit base and The number of elements in ; Step 3: Calculate the weight of each evaluation indicator In potential failure analysis, it is crucial to determine the indicator weights. Therefore, in order to determine the degree of influence of each indicator on potential failures, in this step, the following two formulas are used to derive the standard weights of each evaluation indicator; in, represents the standard weight of the jth standard; Step 4: Construct the rough-Z number decision matrix Since the rough-Z number can comprehensively represent the reliability and uncertainty of the evaluation information, the rough-Z number decision matrix is ​​constructed according to the rough-Z number evaluation calculated in the previous article; The rough Z number decision matrix can be expressed as: RZ = (RZ i j) m×n in, represents the roughness-z number evaluation of the ith potential fault under the jth index; Step 5: Select the fault with the greatest risk based on the improved FMEA method In this step, the improved MULTIMOORA method is used to calculate the comprehensive risk value of each potential fault, and then the fault with the greatest risk is selected; (1) Obtain the risk value of each potential failure based on the ratio system analysis method: Among them, y i represents the risk value of the ith potential failure, j = 1, 2, ..., h is a benefit index, j = h+1, h+2, ..., n is a cost index; in this method, y i The larger the value, the greater the potential failure risk. (2) Obtain the risk value of each potential fault based on the reference point method: Among them, r j represents the reference point of the jth indicator, d ij Indicates the deviation between the evaluation value and the reference point, z i is the risk value of the ith potential failure; z i The smaller it is, the greater the potential failure risk is; (3) Obtain the risk value of each potential fault based on the full multiplication form: Among them, μ i represents the risk value of the ith potential failure, μ i The higher the value, the greater the potential failure risk; (4) Calculate the final risk value of each potential failure: Among them, R(i) represents the final risk value of the ith potential failure, r(y i ), r(z i ), r(μ i ) represent the ranking of the i-th potential fault under the three methods respectively; the larger R(i) is, the higher the ranking is and the greater the risk is.

Images