A Bayesian network-based method for identifying weak links in combined heat and power systems

Abstract

The present invention discloses a method for identifying weak links in a combined heat and power system based on a Bayesian network, which belongs to the technical field of energy systems. The method comprises the following steps: establishing a multi-state-coupling logical relationship model; a temporal simulation process of a Bayesian network of a combined heat and power system taking into account the multi-states of coupling elements; and identifying reliability weak links in a multi-state combined heat and power system of coupling elements based on the Bayesian network temporal simulation. The present invention establishes a "multi-state-coupling" logical relationship model that is closer to actual engineering scenarios. A comparative analysis is conducted on the Bayesian network temporal simulation processes of the distribution subsystem and the thermal subsystem with and without considering the multi-states of coupling elements, which more finely depicts the degree of complex coupling correlation between various energy forms and the states of electrical and thermal load points. The improved Bayesian network temporal simulation reasoning algorithm is used to accurately evaluate the multi-state reliability level of the combined heat and power system and precisely identify the weak links in the system reliability.

Description

Technical Field

[0001] The present invention relates to the technical field of energy systems, and in particular to a Bayesian network-based method for identifying weak links in a combined heat and power system. Background Art

[0002] Overexploitation of fossil fuels has led to environmental pollution and energy crises. Driven by the goal of carbon neutrality, the application of combined heat and power (CHAPS) systems has emerged. Combined heat and power systems (CHAPS) couple electricity and heat energy subsystems. Compared to traditional power systems, CHAPS improve system reliability through multi-energy complementarity and cascaded energy utilization. Their reliability indicators provide valuable information for optimized scheduling and operational planning. Therefore, accurately evaluating the reliability of CHP systems and identifying weaknesses is crucial.

[0003] Combined heat and power (CHP) systems are complex and comprise numerous components. Components can be broadly categorized into independent and coupled components based on their structural type. Independent components (hereafter referred to as components) generally have relatively simple structures, such as wiring, pipe networks, and valves. Coupled components, comprised of several independent components working in conjunction, are core equipment for energy conversion, such as combined heat and power (CHP) units. Currently, most component outage models employ a two-state probabilistic model: "normal-faulty." However, in actual operation, when a component fails, the coupled component does not completely fail, but rather exists in one or more intermediate states. While in these intermediate states, it still maintains the ability to perform some functions. Consequently, coupled component failure states vary significantly, and their reliability assessment results are closely related to the fault characteristics. Therefore, it is necessary to establish a multi-state model for coupled components to accurately assess system reliability and identify weaknesses.

[0004] In the multi-state reliability assessment method of the integrated energy system with multi-energy coupling and conversion in CN202210329836.0, a multi-state reliability model of a multi-energy conversion element is established considering multi-energy conversion, reliability aggregation is performed using a clustering algorithm, and an aggregated multi-state reliability model is formed by mapping; a multi-state reliability model of a multi-energy coupling element is established considering multi-energy coupling, and Gaussian approximation and sampling approximation are used to sequentially perform reliability approximation processing to obtain a multi-state approximate reliability model; the aggregated multi-state reliability model and the multi-state approximate reliability model are processed according to a rapid reliability assessment method to obtain the reliability of the integrated energy system and realize reliability assessment. The traditional reliability assessment method is improved in terms of time and accuracy, the calculation time is reduced, and the estimation of system reliability is realized quickly and effectively.

[0005] CN202210850181.1 A thermal system reliability assessment method based on a Bayesian network adopts a time-sharing method under the premise of considering thermal inertia delay. The proposed reasoning algorithm can not only perform forward reasoning to calculate the system reliability index, but also perform reverse reasoning to identify the weak links in the system reliability.

[0006] In summary, existing technical solutions still have the following shortcomings: 1) They fail to accurately characterize the reliability model of coupled components in multiple states. 2) When analyzing thermal systems, they only consider the effects of thermal inertia, failing to analyze the impact of coupled components in multiple states on the thermal load point. 3) They fail to accurately identify weak links in system reliability. Summary of the Invention

[0007] This invention aims to provide a Bayesian network-based method for identifying weak links in combined heat and power (CHP) systems. This method establishes a "multi-state-coupling" logical relationship model, making it applicable to the Bayesian network of a CHP system. By analyzing the logical relationships between independent and coupled components, and between coupled components and load points within a minimum time period, the normal fault states and durations of electrical and thermal load points are analyzed, providing guidance for system planning and operation.

[0008] To achieve the above object, the present invention provides a method for identifying weak links in a combined heat and power system based on a Bayesian network, comprising the following steps:

[0009] S1. Determine the correspondence between Bayesian network nodes and CHP system components, and establish a CHP unit reliability assessment model based on multi-state-coupling logic relationships;

[0010] S2. Based on the correspondence between the Bayesian network nodes and the cogeneration system components obtained in step S1, a temporal simulation process of the Bayesian network of the cogeneration system taking into account the multi-state of the coupling components;

[0011] S3. According to the timing simulation results obtained in step S2, reliability weak links of the coupled element multi-state cogeneration system based on Bayesian network timing simulation are identified.

[0012] Preferably, in step S1, the Bayesian network nodes include coupling element nodes, element nodes, load nodes, and system nodes; the cogeneration system elements include cogeneration units, lines, pipelines, circuit breakers, valves, electrical load points, and thermal load points.

[0013] Preferably, in step S2, the timing simulation process of the Bayesian network of the combined heat and power system taking into account the multiple states of the coupling elements includes the timing simulation process of the Bayesian network of the coupling elements taking into account the multiple states, the timing simulation process of the Bayesian network of the power distribution subsystem taking into account the multiple states of the coupling elements, and the timing simulation process of the Bayesian network of the thermal subsystem taking into account the multiple states of the coupling elements.

[0014] Preferably, the temporal simulation process of the Bayesian network of coupled elements taking into account multiple states comprises the following steps:

[0015] Assume that the main components of the cogeneration unit are two-state models, which alternate between normal and fault states during the simulation period.

[0016] Assuming that the state duration of each component obeys the exponential distribution, the normal state duration t of each component in the simulation time period is obtained by formulas (1) and (2): n and the duration of the fault state (repair time) t r , where formula (1) and formula (2) are as follows:

[0017]

[0018] In formula (1) and (2), f0(t) and f1(t) are random numbers uniformly distributed in the interval (0, 1), MTTF is the mean time to failure, and MTTR is the mean repair time of the component;

[0019] Compare the state duration of each component and find the minimum value as the minimum duration D min1 .

[0020] Preferably, the Bayesian network temporal simulation process of the power distribution subsystem taking into account the multi-states of the coupling elements comprises the following steps:

[0021] Assuming that each component in the distribution subsystem is a two-state model, the normal and fault duration of each component in the distribution subsystem are obtained by formulas (1) and (2);

[0022] According to the multi-state model of the coupling element cogeneration element obtained in step S1, the state duration of each element is compared to find a new minimum value as the minimum duration D min2 .

[0023] Preferably, the temporal simulation process of the Bayesian network of the thermal subsystem taking into account the multi-states of the coupling elements comprises the following steps:

[0024] Assuming that each component in the thermal system is a two-state model, the normal and fault duration of each component are obtained by formulas (1) and (2) and thermal inertia is taken into account;

[0025] Compare the state duration of each component and find the new minimum value as the minimum duration D min3 .

[0026] Preferably, in step S3, the weak link identification method is specifically as follows:

[0027] S301, initializing each data, setting the confidence probability α=95%, and the number of cycles constituting independent generalized events to 20,000;

[0028] S302: Set the components in the cogeneration unit to be two-state models, perform time series simulation on the state of each component, and find the minimum duration D. min1 ,In the minimum time period, the state and duration of the cogeneration unit are obtained according to the “multi-state-coupling” logical relationship;

[0029] S303, assuming that the circuit components in the distribution subsystem are all two-state models, combined with the state and duration of the cogeneration unit, obtain the new minimum duration D min2 For the thermal subsystem, combined with the state and duration of the cogeneration unit, a new minimum duration D is obtained. min3 ;

[0030] S304, accumulating the minimum duration T min According to the causal relationship between the electric and heating load nodes and the system nodes, the fault times and fault durations of the system nodes are analyzed and calculated. min The status and corresponding time in the system are used to obtain the accumulated time of system failure;

[0031] S305: Generate the next state and duration of each component, repeat steps S302 to S305 until the number of cycles exceeds a given value, determine whether the convergence criterion is met, and then end the loop;

[0032] S306, by accumulating several minimum durations T min System failure time T j And the failure time of each component t ij Parameters such as , calculate the conditional probability p of the i-th component failure when the system fails i .

[0033]

[0034] In formula (3), t ij is the failure time of the i-th component in the j-th time period, T j is the simulation time of the jth time period;

[0035] S307: After calculating the conditional probability of failure of all components, compare the conditional probabilities of the faulty components to analyze and identify the weak links in the reliability of the system.

[0036] Therefore, the present invention adopts a Bayesian network-based method for identifying weak links in a combined heat and power system, which has the following beneficial effects:

[0037] This paper establishes a "multi-state-coupling" logical relationship model that more closely resembles actual engineering scenarios, targeting the various operating states of coupling elements in a combined heat and power (CHP) system. A comparative analysis of Bayesian network temporal simulations of the distribution subsystem and thermal subsystem with and without consideration of the multi-state coupling elements is conducted, providing a more detailed depiction of the complex coupling relationships between various energy sources and the states of electrical and thermal load points. Using an improved Bayesian network temporal simulation inference algorithm, the multi-state reliability level of the CHP system is accurately assessed, and weaknesses in system reliability are precisely identified.

[0038] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 This is a schematic diagram of the system flow of a Bayesian network-based method for identifying weak links in a combined heat and power system according to the present invention;

[0040] Figure 2 A schematic diagram of the structure of a combined heat and power (CHP) unit according to a Bayesian network-based method for identifying weak links in a combined heat and power system of the present invention;

[0041] Figure 3 This is a schematic diagram of a "multi-state-coupling" logical relationship model of a Bayesian network-based method for identifying weak links in a combined heat and power system according to the present invention;

[0042] Figure 4 A schematic diagram of a time series simulation process of a Bayesian network of a coupling element taking into account multiple states in a Bayesian network-based method for identifying weak links in a combined heat and power system according to the present invention;

[0043] Figure 5 Schematic diagram of a Bayesian network time series simulation process of a power distribution subsystem taking into account multiple states of coupling elements in a Bayesian network-based method for identifying weak links in a combined heat and power system according to the present invention;

[0044] Figure 6 The figure is a schematic diagram of the Bayesian network time series simulation process of the thermal subsystem taking into account the multi-state and thermal inertia of the coupling elements in the Bayesian network-based weak link identification method of the combined heat and power system of the present invention. DETAILED DESCRIPTION

[0045] The technical solution of the present invention is further described below with reference to the accompanying drawings and embodiments.

[0046] Unless otherwise defined, the technical or scientific terms used in the present invention shall have the usual meanings understood by persons of ordinary skill in the field to which the present invention belongs. The words "first", "second" and similar terms used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. Words such as "include" or "comprise" mean that the elements or objects preceding the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. Words such as "connect" or "connected" are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the object being described changes, the relative positional relationship may also change accordingly.

[0047] Example

[0048] like Figure 1-6 As shown, the present invention provides a method for identifying weak links in a combined heat and power system based on a Bayesian network, comprising the following steps:

[0049] S1. Determine the correspondence between Bayesian network nodes and CHP system components, and establish a CHP unit reliability assessment model based on multi-state-coupling logic relationships;

[0050] S2. Based on the correspondence between the Bayesian network nodes and the cogeneration system components obtained in step S1, a temporal simulation process of the Bayesian network of the cogeneration system taking into account the multi-state of the coupling components;

[0051] S3. According to the timing simulation results obtained in step S2, reliability weak links of the coupled element multi-state cogeneration system based on Bayesian network timing simulation are identified.

[0052] In step S1, the Bayesian network nodes include coupling element nodes, element nodes, load nodes, and system nodes; the cogeneration system elements include cogeneration units, lines, pipelines, circuit breakers, valves, electrical load points, and thermal load points. The corresponding relationship between the Bayesian network nodes and the cogeneration system elements is shown in Table 1.

[0053] Table 1 Correspondence between Bayesian network nodes and cogeneration system components

[0054]

[0055] Combined heat and power (CHP) unit mainly consists of internal combustion engine, cylinder jacket water heat exchanger, flue gas heat exchanger, biogas boiler and other components, such as Figure 2 The working principle can be summarized as follows: First, biomass feedstock is burned and fermented to produce biogas. A portion of the biogas is mixed with air and fed into the cylinder, where it is compressed and ignited. The ignited compressed gas expands, pushing the piston back and forth, thereby driving the internal combustion engine to generate electricity and power the electrical load. The remaining biogas is burned in the biogas boiler to generate heat. While the internal combustion engine generates electricity, it also produces high-temperature flue gas at around 500°C and jacket water at around 80°C. The waste heat is recovered through the flue gas heat exchanger and the jacket water heat exchanger, respectively. The heat from the high-temperature flue gas is used to dry the biogas residue. The heat from the jacket water heat exchanger and the heat generated by the biogas boiler are combined and fed into the hot water storage tank to keep the anaerobic fermentation tank warm, helping to increase biogas production.

[0056] The Bayesian network logical relationship model of coupling elements is as follows:

[0057] Suppose A, B, C... in the coupling element represent the node variables of each element, and D represents the node variable of the coupling element. Each element node has two states: normal and faulty, where 0 represents the faulty state and 1 represents the normal state. The components inside the coupling element work independently. When one component fails, the other components may not fail. Therefore, the coupling element can have three or more states, where 0 represents the faulty state, 1 represents the normal state, and the others each represent an intermediate state. Unlike traditional two-state elements, when the coupling element is in an intermediate state, the coupling element is not completely faulty, but still has the function of completing a certain part of the function. Therefore, the relationship between each component in the coupling element and the coupling element is named the "multi-state-coupling" logical relationship, such as Figure 3 shown.

[0058] The establishment of the reliability assessment model of the cogeneration unit is as follows:

[0059] Corresponding to Figure 2 The structure of the cogeneration unit is analyzed, and the working status of the cogeneration unit is analyzed: when all components are normal and there is no fault, the cogeneration unit is in the "normal operation" state, and the state is 1; when the internal combustion engine is normal, and the cylinder jacket water heat exchanger is in a faulty state, the cylinder jacket water heat exchanger cannot work normally to reduce the temperature of the internal combustion engine, so the internal combustion engine is reduced to the predetermined output, and the flue gas heat exchanger is also affected by it and reduces the heat production. At this time, the cogeneration unit is in the "derated power supply" state, and the state is 2; when the internal combustion engine and the cylinder jacket water heat exchanger are in a faulty state, the cylinder jacket water heat exchanger cannot work normally to reduce the temperature of the internal combustion engine, so the internal combustion engine is reduced to the predetermined output, and the flue gas heat exchanger is also affected by it and reduces the heat production. At this time, the cogeneration unit is in the "derated power supply" state, and the state is 2; when the internal combustion engine and the cylinder jacket water heat exchanger are in a faulty state, the cylinder jacket water heat exchanger is in a faulty state, and ... When the heat exchanger is in normal condition, the internal combustion engine can supply power normally. If either the flue gas heat exchanger or the biogas boiler is in a faulty state, it will not be able to provide sufficient heat for the load. At this time, the cogeneration unit is in the "reduced heat supply" state, and the state is 3. When the internal combustion engine fails, the jacket water heat exchanger and the flue gas heat exchanger cannot generate heat energy. Therefore, regardless of whether the biogas boiler is working normally, the cogeneration unit cannot supply power or heat normally. At this time, the cogeneration unit is in the "fault shutdown" state, and the state is 0.

[0060] From the above analysis, it can be seen that the four main components in the cogeneration unit exist independently. Therefore, it is difficult to describe the state of the cogeneration unit using the previous "normal-fault" two-state model. Therefore, based on the above "multi-state-coupling" logical relationship, a conditional probability table of the cogeneration unit is established, see Table 2.

[0061] Table 2 Conditional probability table of “multi-state-coupling” logical relationship of cogeneration unit

[0062]

[0063] In step S2, the temporal simulation process of the Bayesian network for the CHP system, which accounts for the multi-state coupling element, includes the temporal simulation of the Bayesian network for the multi-state coupling element, the Bayesian network for the power distribution subsystem, and the Bayesian network for the thermal subsystem. During the temporal simulation, assuming sufficient biomass feedstock and biogas supply, the state of the CHP unit is first determined based on the "multi-state-coupling" logical relationship. Then, based on the energy supply relationship, the state of the electrical and thermal load within the system is determined by combining the states of the components of the power distribution subsystem and the thermal subsystem. This allows the reliability of the entire CHP system to be assessed and weaknesses to be identified.

[0064] (1) The temporal simulation process of the Bayesian network of coupled elements taking into account multiple states includes the following steps:

[0065] Assume that the main components of the cogeneration unit are two-state models, which alternate between normal and fault states during the simulation period.

[0066] Assuming that the state duration of each component obeys the exponential distribution, the normal state duration t of each component in the simulation time period is obtained by formulas (1) and (2): n and the duration of the fault state (repair time) t r , where formula (1) and formula (2) are as follows:

[0067]

[0068] In formula (1) and (2), f0(t) and f1(t) are random numbers uniformly distributed in the interval (0, 1), MTTF is the mean time to failure, and MTTR is the mean repair time of the component; Figure 4 As shown, it indicates the status and duration of the cogeneration unit in different states.

[0069] Compare the state duration of each component and find the minimum value as the minimum duration D min1 .

[0070] In the second minimum duration T 2min That is, during the period t1-t2, the internal combustion engine, flue gas heat exchanger, and biogas boiler were all operating normally, but the jacket water heat exchanger was faulty. Since the jacket water heat exchanger was unable to cool the internal combustion engine, the engine was reduced to its predetermined output, reducing power supply. The flue gas heat exchanger also reduced its heat output due to the reduced power supply from the internal combustion engine. Even though the biogas boiler could continue to heat normally, it could not provide sufficient heat. Therefore, the cogeneration unit's state is 2, indicating a reduced power supply state, which shows the number and duration of reduced power supply states.

[0071] The third minimum duration T 3min That is, during the period t2-t3, the internal combustion engine and biogas boiler are operating normally, while the remaining components are faulty. At this point, the internal combustion engine still operates at its designated output, providing some power, while the biogas boiler can still supply heat but cannot meet demand. Therefore, the cogeneration unit is in state 2, indicating a reduced power supply state. The cumulative number and duration of reduced power supply states are shown.

[0072] In the fourth minimum duration period T 4min That is, during the period t3-t4, the internal combustion engine, the jacket water heat exchanger, and the flue gas heat exchanger are all operating normally, while the biogas boiler is faulty. At this time, the internal combustion engine is operating normally, and the jacket water heat exchanger and flue gas heat exchanger are also providing heat normally. The biogas boiler is unable to provide heat due to the fault, but it can still supply sufficient heat. Therefore, the cogeneration unit is in state 3, indicating reduced heat supply. This shows the cumulative number and duration of reduced heat supply states.

[0073] In the fifth minimum duration period T 5min That is, during the period t4-t5, the internal combustion engine, the jacket water heat exchanger, and the biogas boiler are all functioning normally, but the flue gas heat exchanger is faulty. At this time, the internal combustion engine can provide power normally, and the jacket water heat exchanger and biogas boiler are functioning normally and providing sufficient heat. Therefore, the cogeneration unit is in state 3, which indicates reduced heat supply. The cumulative number and duration of reduced heat supply states are shown.

[0074] In the sixth minimum duration period T 6min , meaning that during the period t5-t6, the internal combustion engine failed, while all other components were normal. At this time, the internal combustion engine could not provide power, and the jacket water heat exchanger and flue gas heat exchanger could not provide heat. Although the biogas boiler could function normally, it could not provide sufficient heat. Therefore, the cogeneration unit's status is 0, indicating a faulty shutdown. The number of fault states and their duration are accumulated.

[0075] In the seventh minimum duration period T 7min , that is, during the period t6-t7, all components are in fault state. Therefore, the state of the cogeneration unit is 0, that is, fault shutdown, and the cumulative number of fault states and duration.

[0076] (2) The Bayesian network temporal simulation process of the distribution subsystem taking into account the multi-state of the coupling element includes the following steps:

[0077] Assuming that each component in the distribution subsystem is a two-state model, the normal and fault duration of each component in the distribution subsystem are obtained by formulas (1) and (2);

[0078] According to the multi-state model of the coupling element cogeneration element obtained in step S1, the state duration of each element is compared to find a new minimum value as the minimum duration D min2 .

[0079] like Figure 5 As shown, in the first minimum duration period T′ 1min , that is, during the period 0-t1, the cogeneration unit and the power distribution subsystem circuits and other components are in normal state. Therefore, the state of the load point is normal, and the number and duration of normal state are accumulated.

[0080] In the second minimum duration T′ 2min , that is, during the period t1-t2, the cogeneration unit is operating normally, but the connected line 2 is faulty, and the load point is unable to receive power. Therefore, the load point is in a faulty state, and the number of fault states and their duration are accumulated.

[0081] In the third minimum duration T′ 3min , that is, during the period t2-t3, the cogeneration unit is in a derated power supply state, and the lines are all in normal state. At this time, it is assumed that the power supply of the cogeneration unit is P S , the power demand of the load point in the microgrid is P D .

[0082] If P S ≥P D , then the electrical load points in the microgrid are all in normal state; if P S <P D At this time, load reduction should be carried out according to the load reduction strategy based on the importance of the electric load point, thereby reducing the pressure on the load point in the microgrid to ensure the stability of the system. Here, we assume that the cogeneration unit can still meet the load point demand, that is, P S ≥P D , so the state of the electric load point is normal, and the number of normal state times and duration are accumulated.

[0083] In the fourth minimum duration period T′ 4min, meaning that during the period t3-t4, the CHP unit was operating at a reduced power supply. However, Line 2 was faulty. Even though the CHP unit could still supply some power, the load point was unable to receive power. Therefore, the load point is in a faulty state, and the number and duration of fault states are accumulated.

[0084] In the fifth minimum duration period T′ 5min , meaning that during the period t4-t5, the CHP unit was in a reduced-heat power supply state. At this time, the CHP unit was supplying power normally, and all circuit components were in a normal state. Therefore, the load point was in a normal state, and the number and duration of normal state events were accumulated.

[0085] In the sixth minimum duration period T′ 6min , meaning that during the period t5-t6, the CHP unit is in a reduced power supply state. While the CHP unit's power supply remains normal, a fault exists in the line. Therefore, the load point is in a fault state, and the number of fault states and their duration are accumulated.

[0086] In the 7th and 8th minimum duration periods T′ 7min , T′ 8min That is, during the periods t6-t7 and t7-t, the cogeneration unit is in a fault shutdown state. At this time, the cogeneration unit cannot supply power. Regardless of the state of the line, the electric load point is in a fault state, and the number of fault states and duration are accumulated.

[0087] The nth minimum duration T′ nmin , that is, t n-1 -t n During the time period, the state of the electric load point is determined based on the state of the cogeneration unit and the line state.

[0088] The above is the temporal simulation process of the Bayesian network of the distribution subsystem considering the multi-state of the coupling element. In order to facilitate the subsequent comparison and analysis of the reliability indicators of the system, Figure 5 The state of the electric load point when the multiple states of the coupling elements are not considered is also analyzed: that is, the electric load point is in normal state only when the cogeneration unit is in normal state and the circuit and other components are in normal state.

[0089] Depend on Figure 5 It can be seen that the load point assessment results differ between those that consider the multi-state coupling element and those that do not. The load point that considers multi-states remains normal even when the coupled cogeneration unit is operating in derated or reduced heat supply states, i.e., during the t2-t3 and t4-t5 periods. The load point that does not consider multi-states remains normal only when the cogeneration unit is operating normally and all other components are operating normally, i.e., during the 0-t1 period. Therefore, the Bayesian network time series simulation analysis of the distribution subsystem using the model that considers multi-state coupling elements is closer to engineering practice.

[0090] (3) The temporal simulation process of the Bayesian network of the thermal subsystem taking into account the multi-state of the coupling elements includes the following steps:

[0091] Assuming that each component in the thermal system is a two-state model, the normal and fault duration of each component are obtained by formulas (1) and (2) and thermal inertia is taken into account;

[0092] Compare the state duration of each component and find the new minimum value as the minimum duration D min3 .

[0093] like Figure 6 As shown, in the first minimum duration period T″ 1min , that is, during the period 0-t1, all components are in normal state and the state of the cogeneration unit is in normal operation, so the state of the heat load point is normal. Due to the insulation performance of the building on the indoor temperature, the temperature of the load point increases from the initial temperature T in It starts to rise gradually until it reaches the maximum temperature T at time t0 max And keep it up, accumulating the number of normal states and their duration.

[0094] In the second minimum duration T″ 2min , that is, during the period t1-t2, at time t1, pipe 2 fails. Even if the cogeneration unit can supply heat normally, the heat load point cannot get heat. Since there is still water in the failed pipe, the temperature can be maintained at T for a short time. max unchanged, define the holding time as t m ,exist Figure 6 Expressed as the time period t1-t1′.

[0095] At t1′, there is no water in the faulty pipe, and the temperature at the load point begins to decrease due to the influence of the external environment, until it drops to T at t1″. m , T m It is defined as the minimum standard temperature for central heating in winter, which is determined by T max Reduce to T m The time of temperature change is defined as t v , expressed as the period t1′-t1″.

[0096] The pipeline was not repaired afterwards, so the temperature continued to drop.

[0097] Compare the heat load point temperature during this period with the minimum standard temperature T for central heating in winter m The size of the thermal load point determines the state of the thermal load point.

[0098] During the period t1-t″1, the temperature of the heat load point is higher than T m, so the heat load point is in normal state, the cumulative normal times and duration; in the period t1″-t2, the temperature of the heat load point is lower than T m , the thermal load point is in fault state, the cumulative number of faults and duration.

[0099] Analyze the third minimum duration T″ 3min , that is, the period t2-t3. At t2, the pipeline fault is repaired and the cogeneration unit is in a derated power supply state. The heat load point can get heat supply, but it cannot meet its demand. Therefore, in the period t2-t′2, the temperature of hot water does not drop after flowing through the pipeline and remains at T h During the period t2-t3, the temperature of the heat load point is always lower than T m , so the thermal load point is in a fault state, and the cumulative number of faults and duration are calculated.

[0100] Analyze the fourth minimum duration period T″ 4min , i.e., the period t3-t4. At t3, the cogeneration unit is in a derated power supply state and can supply some heat, but the pipeline fails again, so the heat load point cannot get heat. During the period t3-t′3, there is still water supply in the faulty pipeline. Then, at t′3, the temperature begins to drop, and finally drops to T at t3″. out And keep it up, T out Indicates the outdoor ambient temperature. During the period t3-t4, the temperature is always lower than T m , so the thermal load point is in a fault state, and the cumulative number of faults and duration are calculated.

[0101] Analyze the fifth minimum duration period T″ 5min , that is, the period from t4 to t5. At t4, the pipeline fault is repaired, and the cogeneration unit is in a reduced heat supply state, which can supply some heat and meet the demand of the heat load point. Therefore, in the period from t4 to t′4, after the hot water flows through the pipeline, the temperature begins to rise. At t″4, the temperature reaches T m , then the temperature continues to rise and reaches T at t″′4 c , because the cogeneration unit can only supply part of the heat, the temperature cannot continue to rise and be maintained.

[0102] During the period t4-t″4, the temperature of the heat load point is lower than T m , so the heat load point is in fault state, the cumulative fault times and duration; in the period t″4-t5, the temperature of the heat load point is higher than T m , so the heat load point is in normal state, and the cumulative normal times and duration.

[0103] Analyze the sixth minimum duration period T″ 6min, that is, the period from t5 to t6. At t5, the pipeline fails again. During the period from t5 to t′5, there is still water in the failed pipeline. Therefore, the temperature starts to drop at t′5 and drops to T at t″5. m , and then the temperature continued to drop.

[0104] During the period of t5-t″5, the temperature of the heat load point is higher than T m , so the heat load point is in normal state, the cumulative normal times and duration; in the period of t″5-t6, the temperature of the heat load point is lower than T m , so the thermal load point is in a fault state, and the cumulative number of faults and duration are calculated.

[0105] Analyze the 7th minimum duration period T″ 7min , that is, the period from t6 to t7. The pipeline fault is repaired, but the cogeneration unit is in a faulty shutdown state and cannot provide heat, so the temperature continues to drop. At t′6, the temperature drops to T out During the period t6-t7, the heat load point is in a fault state, and the number of faults and duration are accumulated.

[0106] Analyze the 8th minimum duration period T″ 8min , that is, the period from t7 to t8. The pipeline fails and the cogeneration unit is in a fault shutdown state, so there is no heat supply and the temperature continues to remain at T out During the period t7-t8, the thermal load point is in a fault state, and the cumulative number of faults and duration are shown.

[0107] The nth minimum duration T″ nmin , that is, t n-1 -t n During the time period, the status of the heat load point is determined based on the status of the cogeneration unit and the pipeline.

[0108] The above is the temporal simulation process of the Bayesian network of the thermal system considering the existence of multiple states of the coupling elements. In order to facilitate the subsequent comparative analysis of the reliability indicators of the system, Figure 6 The state of the thermal load point without considering the multi-state of the coupling element is also analyzed: in the period 0-t1, the cogeneration unit is in normal operation and the pipelines are normal, so the thermal load point is in a normal state; in the period t1-t1′, under the influence of thermal inertia, the thermal load point is in a normal state; after that, the cogeneration unit is in an abnormal state and the thermal load point is in a fault state.

[0109] Depend on Figure 6It can be seen that the load point assessment results differ between those that consider the multiple states of the coupling elements and those that do not. The thermal load point that considers multiple states remains normal when the coupled element cogeneration unit is in a reduced heat supply state, i.e., during the period t″4-t″5. The thermal load point that does not consider multiple states remains normal only when all components are normal, i.e., during the period 0-t″1. Therefore, the model that considers the multiple states of the coupling elements is closer to engineering practice in Bayesian network time-series simulation analysis of thermal systems.

[0110] like Figure 1 As shown, in step S3, the weak link identification method is specifically as follows:

[0111] S301, initializing each data, setting the confidence probability α=95%, and the number of cycles constituting independent generalized events to 20,000;

[0112] S302: Set the components in the cogeneration unit to be two-state models, perform time series simulation on the state of each component, and find the minimum duration D. min1 ,In the minimum time period, the state and duration of the cogeneration unit are obtained according to the “multi-state-coupling” logical relationship;

[0113] S303, assuming that the circuit components in the distribution subsystem are all two-state models, combined with the state and duration of the cogeneration unit, obtain the new minimum duration D min2 For the thermal subsystem, combined with the state and duration of the cogeneration unit, a new minimum duration D is obtained. min3 For each minimum time period, the state and duration of the output side load node are obtained according to the Bayesian network timing simulation process of the distribution subsystem, and the parameters such as the number of faults, fault time, and total simulation time of the load point are accumulated respectively.

[0114] Similarly, for the thermal subsystem, combined with the state and duration of the cogeneration unit, a new minimum duration D is obtained. min3 For each minimum time period, the state and duration of the heat load node are obtained according to the Bayesian network time series simulation process of the thermal subsystem, and the relevant parameters of the load point are also accumulated and recorded.

[0115] S304, accumulating the minimum duration T min According to the causal relationship between the electric and heating load nodes and the system nodes, the fault times and fault durations of the system nodes are analyzed and calculated. min The status and corresponding time in the system are used to obtain the accumulated time of system failure;

[0116] S305: Generate the next state and duration of each component, repeat steps S302 to S305 until the number of cycles exceeds a given value, determine whether the convergence criterion is met, and then end the loop;

[0117] S306, by accumulating several minimum durations T min System failure time T j And the failure time of each component t ij Parameters such as , calculate the conditional probability p of the i-th component failure when the system fails i .

[0118]

[0119] In formula (3), t ij is the failure time of the i-th component in the j-th time period, T j is the simulation time of the jth time period;

[0120] S307: After calculating the conditional probability of failure of all components, compare the conditional probabilities of the faulty components to analyze and identify the weak links in the reliability of the system.

[0121] Therefore, the present invention employs the aforementioned Bayesian network-based method for identifying weak links in cogeneration systems, establishing a "multi-state-coupling" logical relationship model that more closely resembles actual engineering scenarios. A comparative analysis of Bayesian network temporal simulations of the distribution subsystem and thermal subsystem with and without consideration of the multi-states of coupling elements was conducted, providing a more detailed depiction of the complex coupling relationships between various energy sources and the states of electrical and thermal load points. Using an improved Bayesian network temporal simulation inference algorithm, the multi-state reliability level of the cogeneration system was accurately assessed, and weak links in system reliability were precisely identified.

[0122] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit the same. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that they can still modify or replace the technical solutions of the present invention with equivalents, and these modifications or equivalent replacements cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A Bayesian network-based method for identifying weak links in a combined heat and power system, characterized by: The following steps are involved: S1. Determine the correspondence between Bayesian network nodes and CHP system components, and establish a CHP unit reliability assessment model based on multi-state-coupling logic relationships; S2. Based on the correspondence between the Bayesian network nodes and the cogeneration system components obtained in step S1, a temporal simulation process of the cogeneration system Bayesian network taking into account the multi-states of the coupling components; S3. Identify reliability weaknesses of the coupled element multi-state cogeneration system based on Bayesian network timing simulation according to the timing simulation results obtained in step S2; In step S1, the Bayesian network nodes include coupling element nodes, element nodes, load nodes, and system nodes; the cogeneration system elements include cogeneration units, lines, pipelines, circuit breakers, valves, electrical load points, and thermal load points; In step S2, the temporal simulation process of the Bayesian network of the combined heat and power system taking into account multiple states of the coupling elements includes the temporal simulation process of the Bayesian network of the coupling elements taking into account multiple states, the temporal simulation process of the Bayesian network of the power distribution subsystem taking into account multiple states of the coupling elements, and the temporal simulation process of the Bayesian network of the thermal subsystem taking into account multiple states of the coupling elements; In step S3, the weak link identification method is specifically as follows: S301, initialization of each data, setting confidence probability , the number of cycles constituting independent generalized events is 20,000; S302: Set the components in the cogeneration unit to be two-state models, perform time series simulation on the state of each component, and find the minimum duration. , within the minimum time period, the state and duration of the cogeneration unit are obtained according to the "multi-state-coupling" logical relationship; S303, assuming that the circuit components in the distribution subsystem are all two-state models, combined with the state and duration of the cogeneration unit, obtain a new minimum duration , for the thermal subsystem, combined with the state and duration of the cogeneration unit, a new minimum duration is obtained ; S304: Accumulate the minimum durations The number of faults and the time of faults within the time period are analyzed and calculated based on the causal relationship between the electric and heating load nodes and the system nodes. The status and corresponding time in the system are used to obtain the accumulated time of system failure; S305: Generate the next state and duration of each component, repeat steps S302 to S305 until the number of cycles exceeds a given value, determine whether the convergence criterion is met, and then end the loop; S306, by accumulating several minimum durations System failure time within And the failure time of each component Parameters, calculate the conditional probability of failure of the i-th component when the system fails ; （3） In formula (3), For the The element in The failure time within a period of time, For the The simulation time of each time period; S307: After calculating the conditional probability of failure of all components, compare the conditional probabilities of the faulty components to analyze and identify the weak links in the reliability of the system.

2. The Bayesian network-based method for identifying weak links in a combined heat and power system according to claim 1, characterized in that: The temporal simulation process of the Bayesian network of coupled elements taking into account multiple states includes the following steps: Assume that all components in the cogeneration unit are two-state models, and they alternate between normal and fault states during the simulation period. Assuming that the state duration of each component obeys the exponential distribution, the normal state duration of each component in the simulation time period is obtained by formula (1) and (2): and fault state duration and repair time , where formula (1) and formula (2) are as follows: （1） （2） In formula (1) and (2), and is a random number uniformly distributed in the interval (0, 1). The mean time between failures is is the mean repair time of the component; Compare the state duration of each component and find the minimum value as the minimum duration .

3. The Bayesian network-based method for identifying weak links in a combined heat and power system according to claim 2, characterized in that: Bayesian network temporal simulation process of distribution subsystem considering multiple states of coupling elements, The following steps are involved: Assuming that each component in the distribution subsystem is a two-state model, the normal and fault duration of each component in the distribution subsystem are obtained by formulas (1) and (2); According to the multi-state model of the coupling element cogeneration element obtained in step S1, the state duration of each element is compared to find a new minimum value as the minimum duration .

4. The Bayesian network-based method for identifying weak links in a combined heat and power system according to claim 3, characterized in that: The temporal simulation process of the Bayesian network of the thermal subsystem taking into account the multi-state of the coupled elements includes the following steps: Assuming that each component in the thermal system is a two-state model, the normal and fault duration of each component are obtained by formulas (1) and (2) and thermal inertia is taken into account; Compare the state duration of each component and find the new minimum value as the minimum duration .

Images