A method for predicting solar heat and power failure based on neural network model
Through the failure prediction method based on neural network model, the potential faults of the solar combined heat and power supply system are identified in advance, and the scope of the failure impact is determined, which solves the risk of emergency repairs by maintenance personnel in plateau areas in severe weather, and realizes a safe and efficient maintenance plan, reducing the risk of maintenance. Job risks and losses.
Patent Information
- Application Number
- CN202411231503.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-04
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2044-09-04
AI Technical Summary
In plateau areas, the installation and maintenance of the solar combined heat and power supply system is difficult and there are high dangers in bad weather, which leads to maintenance personnel facing the risk of sudden failure and emergency repairs, hindering the promotion and application of the system.
The fault prediction method based on neural network model is adopted to obtain and train historical operation data, establish a fault prediction model, warning of potential faults in advance, determine the scope of the fault impact, optimize maintenance plans, and avoid emergency repairs in severe weather.
Effectively warning of potential failures in advance, reduce the work risks of maintenance personnel, ensure personal safety, improve maintenance efficiency, and reduce losses.
Smart Images

Figure CN118735282B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of fault prediction, and in particular to a solar heat and power cogeneration fault prediction method based on a neural network model. Background Art
[0002] Solar cogeneration technology is widely used in various regions, especially in plateau areas, which effectively alleviates the local energy supply pressure. Unlike other regions, in plateau areas, due to the particularity of the natural environment, the actual layout and maintenance of solar cogeneration systems are very difficult, and the corresponding installation and maintenance personnel face greater risks, which will be even worse in bad weather and extreme weather. This has hindered the further promotion and application of solar cogeneration systems.
[0003] In view of this, this application is hereby filed. Summary of the invention
[0004] The purpose of the present invention is to provide a solar cogeneration fault prediction method based on a neural network model, which can effectively provide early warning of potential failure risks in system operation, so that maintenance personnel can reasonably plan maintenance time in advance, thereby avoiding bad weather, reducing work risks, and avoiding maintenance personnel from performing emergency repairs in bad weather environments due to sudden failures, which is beneficial to maintaining the personal safety of maintenance personnel.
[0005] The embodiment of the present invention is achieved as follows:
[0006] A solar heat and power cogeneration fault prediction method based on a neural network model, comprising:
[0007] A historical operation data set of the solar thermal power cogeneration system is obtained, and actual fault values and actual fault locations are attached to the data in the historical operation data set.
[0008] A fault prediction neural network model is established for the solar cogeneration system, and the training of the fault prediction neural network model is completed based on the historical operation data set with actual fault values and actual fault locations.
[0009] The current operating data set of the solar thermal power cogeneration system is obtained, and the current operating data set is input into the fault prediction neural network model to obtain the corresponding predicted fault value and predicted fault location. If the predicted fault value is greater than the fault threshold, it is determined that the functional module corresponding to the predicted fault location has a fault risk.
[0010] Determine the dependency structure between the various functional modules of the solar heat and power cogeneration system, the dependency structure includes mutual dependence and influence coefficient, and determine the fault impact range of the functional modules with fault risks based on the dependency structure, the fault impact range includes fault propagation path and impact degree.
[0011] Furthermore, the current operating data set is obtained by the monitoring element, each functional module is configured with a data transmission module, and all monitoring elements of the same functional module are signal-connected to its data transmission module.
[0012] A data transmission line is set up, and the data transmission modules of all functional modules are connected to the data transmission line and send their respective current operation data sets to the control center through the data transmission line. On the data transmission line, the data transmission modules are in a series relationship.
[0013] A rule table is set, which includes an arrangement order of each functional module. The arrangement order is the same as the arrangement order of each functional module on the data transmission line. In the arrangement order, the functional module located at the farthest end of the data transmission line is located first.
[0014] In the rule table, an identifier set is set for each functional module, each identifier in each identifier set is configured with a corresponding time point, and the identifiers in the identifier set of each functional module are in a one-to-one correspondence. The monitoring element obtains the current running data set according to the time point in the identifier set.
[0015] The data transmission module located at the farthest end of the data transmission line adds an identifier of the corresponding time point at the end of the data packet when sending the current running data set.
[0016] When the data transmission module located downstream of the data transmission line receives the upstream data packet, it confirms whether the identifier at the end of the data packet is the identifier of the previous function module of the current function module at the corresponding time point in the rule table. If so, the data packet of the current running data set of the current function module is attached to the end of the data packet, and the identifier of the time point corresponding to the current function module is added to the end of the newly added data packet. If not, the received upstream data packet is directly sent downstream.
[0017] Furthermore, each functional module is sorted according to the failure rate. On the data transmission line, starting from the far end of the data transmission line, the data transmission modules of the functional modules are arranged in descending order of failure rate until they are arranged to the near end of the data transmission line.
[0018] Furthermore, when the control center receives a data packet containing the current running data set sent by the data transmission line, it confirms the identifier at the end of the data packet and marks the identifier in the rule table, and at the same time marks the identifiers of all upstream functional modules of the functional module corresponding to the identifier at the same time point.
[0019] Furthermore, the rule table is updated regularly.
[0020] The beneficial effects of the technical solution of the embodiment of the present invention include:
[0021] During the operation of the solar energy cogeneration fault prediction method based on the neural network model provided by the embodiment of the present invention, the solar energy cogeneration fault prediction method based on the neural network model can mark the location where the fault risk exists, and can also determine the actual fault impact range of the fault according to the dependency structure.
[0022] When multiple potential fault points occur at the same time, maintenance personnel can determine the maintenance priority based on the fault locations and the corresponding fault impact ranges, thereby minimizing possible losses.
[0023] In general, the solar cogeneration fault prediction method based on the neural network model provided in the embodiment of the present invention can effectively provide early warning of potential failure risks in system operation, so that maintenance personnel can reasonably plan maintenance time in advance, thereby avoiding bad weather, reducing work risks, and avoiding maintenance personnel from performing emergency repairs in bad weather environments due to sudden failures, which is beneficial to maintaining the personal safety of maintenance personnel. BRIEF DESCRIPTION OF THE DRAWINGS
[0024] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.
[0025] Figure 1 A schematic diagram of the data transmission module of each functional module connected to the data transmission line;
[0026] Figure 2 This is a schematic diagram of the composition of the rule table;
[0027] Figure 3 It is a schematic diagram of a data packet passing through the data transmission module 2 (when the identifier matches);
[0028] Figure 4 It is a schematic diagram of a data packet passing through the data transmission module 2 (when the identifier does not match);
[0029] Figure 5 Diagram after marking the rule table for the control center. DETAILED DESCRIPTION
[0030] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.
[0031] Therefore, 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 invention claimed for protection, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0032] The flowcharts used in this specification are used to illustrate the operations performed by the system according to the embodiments of this specification. It is understood that the operations of each step are not necessarily performed precisely in order. On the contrary, the steps can be processed in reverse order or simultaneously. At the same time, other operations can also be added to these processes, or one or more operations can be removed from these processes.
[0033] This embodiment provides a solar heat and power cogeneration fault prediction method based on a neural network model, which includes the following steps:
[0034] S1. Obtain a historical operation data set of the solar thermal power cogeneration system, and attach an actual fault value and an actual fault location to the data in the historical operation data set.
[0035] S2. Establish a fault prediction neural network model for the solar thermal power cogeneration system, and complete the training of the fault prediction neural network model based on the historical operation data set with actual fault values and actual fault locations.
[0036] S3. Obtain the current operating data set of the solar thermal power cogeneration system, and input the current operating data set into the fault prediction neural network model to obtain the corresponding predicted fault value and predicted fault position. If the predicted fault value is greater than the fault threshold, it is determined that the functional module corresponding to the predicted fault position has a fault risk.
[0037] S4. Determine the dependency structure between the functional modules of the solar thermal power cogeneration system, which includes the interdependence and the influence coefficient, and determine the fault impact range of the functional modules with fault risks based on the dependency structure, which includes the fault propagation path and the impact degree. Among them, the influence coefficient refers to the degree to which other functional modules with interdependence are affected by a fault when a functional module fails.
[0038] During operation, the solar cogeneration fault prediction method based on the neural network model can mark the locations where there are fault risks, and can also determine the actual fault impact range of the fault according to the dependency structure.
[0039] When multiple potential fault points occur at the same time, maintenance personnel can determine the maintenance priority based on the fault locations and the corresponding fault impact ranges, thereby minimizing possible losses.
[0040] In general, the solar cogeneration fault prediction method based on the neural network model provided in this embodiment can effectively provide early warning of potential failure risks in system operation, so that maintenance personnel can reasonably plan maintenance time in advance, thereby avoiding bad weather, reducing work risks, and avoiding maintenance personnel from performing emergency repairs in bad weather environments due to sudden failures, which is beneficial to maintaining the personal safety of maintenance personnel.
[0041] In this embodiment, the current running data set is obtained by the monitoring element, each functional module is configured with a data transmission module, and all monitoring elements of the same functional module are signal-connected to its data transmission module.
[0042] Set up a data transmission line. The data transmission modules of all functional modules are connected to the data transmission line and send their current operation data sets to the control center through the data transmission line. On the data transmission line, the data transmission modules are in series relationship, such as Figure 1 shown.
[0043] A rule table is set, which includes an arrangement order of each functional module. The arrangement order is the same as the arrangement order of each functional module on the data transmission line. In the arrangement order, the functional module located at the farthest end of the data transmission line is located first.
[0044] In the rule table, such as Figure 2 As shown, an identifier set is set for each functional module, each identifier in each identifier set is configured with a corresponding time point, and the identifiers in the identifier set of each functional module are in a one-to-one correspondence. The monitoring element obtains the current running data set according to the time point in the identifier set.
[0045] The data transmission module at the farthest end of the data transmission line adds the identifier of the corresponding time point at the end of the data packet when sending the current running data set, such as Figure 3 shown.
[0046] When the data transmission module located downstream of the data transmission line receives the upstream data packet, it confirms whether the identifier at the end of the data packet is the identifier of the previous function module of the current function module at the corresponding time point in the rule table. If so, the data packet of the current running data set of the current function module is attached to the end of the data packet, and the identifier of the time point corresponding to the current function module is added to the end of the newly added data packet, such as Figure 3 If not, the received upstream data packet is directly sent downstream, as shown in Figure 4 shown.
[0047] In actual use, the rule table can be updated regularly.
[0048] It is understood that monitoring elements include but are not limited to sensors. The historical operation data set and the current operation data set may include monitoring data, operating condition data, related operation data, etc. The monitoring data may include temperature data, pressure data, photothermal efficiency data, photoelectric efficiency data, etc., but are not limited thereto. The data that needs to be collected can be flexibly adjusted according to actual needs.
[0049] In other embodiments of the present invention, each functional module can be sorted according to the failure rate, and on the data transmission line, starting from the far end of the data transmission line, the data transmission modules of the functional modules are arranged in descending order of failure rate until they are arranged to the near end of the data transmission line. With such a design, during use, the current running data set of the functional module with a higher failure rate can be more easily sent to the control center, thereby effectively reducing the failure risk of the functional module with a higher failure rate.
[0050] Back to the present embodiment, when the control center receives a data packet containing the currently running data set sent by the data transmission line, it confirms the identifier at the end of the data packet and marks the identifier in the rule table, and at the same time marks the identifiers of all upstream functional modules of the functional module corresponding to the identifier at the same time point.
[0051] When the control center finally marks the completed rule table as Figure 5 As shown, the marked identifier is enclosed by an ellipse. Figure 5 It can be seen that identifier 2'' and identifier 2''' corresponding to time point 2 are not marked, identifier 3''' corresponding to time point 3 is not marked, identifier 5', identifier 5'', and identifier 5''' corresponding to time point 5 are not marked, and all identifiers corresponding to time point N are not marked.
[0052] That is to say, the data packet corresponding to time point N is lost, there is data corruption between function module three and function module two (corresponding to identifier 2'' and identifier 2''' are not marked), there is data corruption between function module four and function module three (corresponding to identifier 3''' are not marked), and there is data corruption between function module two and function module one (corresponding to identifier 5', identifier 5'', and identifier 5''' are not marked).
[0053] Through this design, the beneficial effects of the solar heat and power cogeneration fault prediction method based on the neural network model provided by the embodiment of the present invention include:
[0054] (1) Save data transmission wires, effectively shorten the total length of the wires, and significantly reduce the probability of signal loss and data errors.
[0055] (2) It can realize data self-checking within the system, effectively ensuring the reliability of the transmission of the current running data set, thereby avoiding system false alarms and improving the accuracy of fault prediction and troubleshooting. It can also prevent maintenance personnel from doing useless work, effectively reducing the work risks of maintenance personnel in plateau areas.
[0056] (3) It has good adaptability to the solar thermal power cogeneration system, which facilitates the more flexible arrangement and installation of the solar thermal power cogeneration system according to the actual situation. To adapt to it, it only needs to adaptively adjust the access sequence of each functional module on the data transmission line and adjust the rule table accordingly, which is simple and convenient.
[0057] (4) The control center can more easily identify the interval where data errors occurred during the data transmission process in combination with the rule table, thereby facilitating the system to implement internal self-inspection and review, and reducing the system's false alarms of faults.
[0058] (5) During the data packet transmission process, each time the data transmission module reaches the next functional module, only the identifier at the end of the data packet will be checked, which has higher execution efficiency and is not likely to cause damage to other parts of the data packet, effectively improving data security and integrity.
[0059] To sum up, the solar cogeneration fault prediction method based on the neural network model provided in the embodiment of the present invention can effectively provide early warning of potential failure risks in system operation, so that maintenance personnel can reasonably plan maintenance time in advance, thereby avoiding bad weather, reducing work risks, and avoiding maintenance personnel from performing emergency repairs in bad weather environments due to sudden failures, which is beneficial to maintaining the personal safety of maintenance personnel.
[0060] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.
Claims
1. A solar heat and power cogeneration fault prediction method based on a neural network model, characterized in that: include: Acquire a historical operation data set of the solar thermal power cogeneration system, and attach an actual fault value and an actual fault location to the data in the historical operation data set; Establishing a fault prediction neural network model for the solar thermal power cogeneration system, and completing the training of the fault prediction neural network model based on the historical operation data set to which the actual fault value and the actual fault location are attached; Acquire a current operating data set of the solar thermal power cogeneration system, input the current operating data set into a fault prediction neural network model to obtain a corresponding predicted fault value and a predicted fault position, and if the predicted fault value is greater than a fault threshold, determine that a functional module corresponding to the predicted fault position has a fault risk; Determine the dependency structure between the functional modules of the solar thermal power cogeneration system, the dependency structure includes mutual dependency and influence coefficient, and determine the fault impact range of the functional module with fault risk according to the dependency structure, the fault impact range includes fault propagation path and impact degree; The current running data set is obtained by the monitoring element, each of the functional modules is configured with a data transmission module, and all monitoring elements of the same functional module are signal-connected to its data transmission module; Setting a data transmission line, wherein the data transmission modules of all the functional modules are connected to the data transmission line and send their respective current operation data sets to the control center through the data transmission line, and the data transmission modules are in a series relationship on the data transmission line; Setting a rule table, wherein the rule table includes an arrangement order of each of the functional modules, wherein the arrangement order is the same as an arrangement order of each of the functional modules on the data transmission line, and in the arrangement order, the functional module located at the farthest end of the data transmission line is located first; In the rule table, an identifier set is set for each functional module, each identifier in each identifier set is configured with a corresponding time point, and there is a one-to-one correspondence between the identifiers in the identifier set of each functional module; the monitoring element obtains the current operation data set according to the time point in the identifier set; The data transmission module located at the farthest end of the data transmission line adds an identifier of the corresponding time point at the end of the data packet when sending the current running data set; When the data transmission module located downstream of the data transmission line receives an upstream data packet, it confirms whether the identifier at the end of the data packet is the identifier of the previous function module of the current function module at the corresponding time point in the rule table; if so, a data packet of the current running data set of the current function module is attached to the end of the data packet, and the identifier of the current function module at the time point corresponding to the end of the newly added data packet is added; if not, the received upstream data packet is directly sent downstream; Sorting each functional module according to the failure rate, and arranging the data transmission modules of the functional modules in descending order of the failure rate starting from the far end of the data transmission line, until arranging them to the near end of the data transmission line; When the control center receives the data packet containing the currently running data set sent by the data transmission line, it confirms the identifier at the end of the data packet and marks the identifier in the rule table, and at the same time marks the identifiers of all upstream functional modules of the functional module corresponding to the identifier at the same time point.
2. The solar heat and power cogeneration fault prediction method based on the neural network model according to claim 1 is characterized in that: The rule table is updated regularly.
