Engine fault intelligent diagnosis system based on Internet of Things

Abstract

The invention relates to the technical field of engine data analysis, in particular to an intelligent engine fault diagnosis system based on the Internet of Things, which comprises an engine data acquisition module used for acquiring engine operation information, engine working condition information and nitrogen oxide concentration in real time; the data node construction module is used for constructing data nodes to obtain node information; the node data analysis module is used for analyzing node data types, and the data extraction and analysis module is used for extracting node information to obtain sample information and analyzing sample node features according to the sample information; the similarity analysis module is used for analyzing the load similarity according to the sample node features; the sample matching analysis module is used for analyzing the node fault matching degree; and the engine fault early warning module is used for carrying out early warning on the engine fault according to the node fault matching degree. According to the invention, accurate diagnosis and early warning of engine faults are realized.

Description

Technical Field

[0001] The present invention relates to the technical field of engine data analysis, and in particular to an intelligent engine fault diagnosis system based on the Internet of Things. Background Art

[0002] Engine fault warning is mainly based on the Internet of Things, sensor technology and data analysis. By real-time monitoring of engine operating parameters and using advanced data mining and machine learning algorithms to identify anomalies, an early warning mechanism is established to ensure that maintenance personnel can obtain early warning information in a timely manner and improve the response speed of fault handling.

[0003] China Patent Publication No.: CN113358360A discloses an engine fault warning system, including: a transport monitoring terminal, a driver application terminal, a fleet management terminal, a transfer center terminal, a maintenance terminal, a supplier detection terminal, and a transport vehicle terminal set in the transport vehicle cockpit. The traditional method of relying on the driver to discover vehicle engine failures during transportation is upgraded to a full-time full-link automatic alarm for the system, and can simultaneously alarm multiple upstream and downstream systems to provide a solution. This invention realizes the fault analysis of engine operation data during vehicle transportation operations, but does not realize the analysis of engine failures during continuous medium and high load operation of the engine. There are problems such as low efficiency in engine data analysis and inaccurate engine failure warnings. Summary of the invention

[0004] The object of the present invention is to provide an intelligent engine fault diagnosis system based on the Internet of Things to solve at least one of the problems existing in the prior art.

[0005] To achieve the above object, the present invention adopts the following technical solutions: An intelligent engine fault diagnosis system based on the Internet of Things, comprising: Engine data acquisition module, used to collect engine operation information, engine operating condition information and nitrogen oxide concentration in real time; Data node construction module, used to construct data nodes to obtain node information; A data extraction and analysis module is used to extract node information according to the node data type and the engine operating condition information to obtain sample information, and analyze the sample node characteristics according to the sample information; the data extraction and analysis module includes a sample node extraction unit and a sample characteristic analysis unit, the sample node extraction unit is used to extract node information according to the node data type and the engine operating condition information to obtain sample information, and the sample characteristic analysis unit is used to analyze the sample node characteristics according to the sample information; A sample matching analysis module is used to analyze the node fault matching degree according to the node information, node data type, engine operating condition information, load similarity, sample information and sample node characteristics; The engine fault warning module is used to warn of engine faults according to the node fault matching degree.

[0006] Furthermore, the system also includes: a node data analysis module, which is used to analyze the main vibration frequency, node-related parameters, node pressure coefficient and node pressure theoretical value according to the node information, and analyze the node data type according to the main vibration frequency, node-related parameters, node pressure coefficient and node pressure theoretical value, and the node data analysis module includes a vibration main frequency analysis unit, a vibration type analysis unit, a temperature current analysis unit, a temperature type analysis unit, a pressure data analysis unit and a pressure type analysis unit.

[0007] Further, the vibration main frequency analysis unit extracts the vibration main frequency from the node vibration frequency using fast Fourier transform; The vibration type analysis unit analyzes the low frequency amplitude, the medium frequency amplitude and the high frequency amplitude according to the amplitude of the frequency points in different frequency bands in the main vibration frequency, and determines the node vibration type according to the low frequency amplitude, the medium frequency amplitude and the high frequency amplitude, and the node vibration type includes one category and two categories; The temperature and current analysis unit analyzes the node-related parameters according to the node cylinder temperature and the node load current to obtain the node-related parameters Q(i); The temperature type analysis unit determines the node temperature type according to the node association parameter and the node nitrogen oxide concentration, and the node temperature type includes type I and type II; The pressure data analysis unit uses the node number corresponding to the cold start time as the cold start node number, and analyzes the node pressure coefficient according to the cold start node number and the node oil pressure to obtain the node pressure coefficient k; The pressure data analysis unit analyzes the node pressure theoretical value according to the node oil pressure, the node speed and the node pressure coefficient to obtain the node pressure theoretical value P1(i); The pressure type analysis unit determines the node pressure type according to the node pressure coefficient and the node pressure theoretical value, and the node pressure type includes one type and two types.

[0008] Further, the sample node extraction unit extracts node information whose engine operating condition information is idle and whose node vibration type, node temperature type and node pressure type are all of the second category as idle sample information; extracts node information whose engine operating condition information is medium load and whose node vibration type, node temperature type and node pressure type are at least of the first category as first medium load sample information; extracts node information whose engine operating condition information is medium load and whose node vibration type, node temperature type and node pressure type are all of the second category as second medium load sample information; extracts node information whose engine operating condition information is high load and whose node vibration type, node temperature type and node pressure type are at least of the first category as first high load sample information; extracts node information whose engine operating condition information is high load and whose node vibration type, node temperature type and node pressure type are all of the second category as second high load sample information.

[0009] Furthermore, the sample feature analysis unit analyzes the sample node feature according to the sample information to obtain the sample node feature H(u), where u represents the sample data number, u∈{1,2,3,4,5}.

[0010] Furthermore, the system further comprises: a similarity analysis module for analyzing the load similarity according to the sample node characteristics; The similarity analysis module analyzes the load similarity according to the sample node feature H(u) to obtain the load similarity S.

[0011] Furthermore, the sample matching analysis module is provided with a matching degree analysis unit, which is used to analyze the node fault matching degree under different engine operating condition information according to the node information and the sample node feature H(u) to obtain the fault matching degree R(i).

[0012] Furthermore, the sample matching analysis module is also provided with a similarity correlation analysis unit, which is used to improve the analysis process of node fault matching according to load similarity, and the improved node fault matching is R1(i).

[0013] Furthermore, the sample matching analysis module is also provided with a type matching analysis unit, which is used to count the number of analysis cycles whose node data types are exactly the same as the node data types of the current analysis cycle after the analysis cycle with node number j as the number of similar cycles when at least one of the node vibration type, node temperature type and node pressure type of the current analysis cycle is of the same type, and improve the analysis process of the load similarity according to the number of similar cycles, and the improved load similarity is S1.

[0014] Furthermore, the engine fault warning module warns of engine fault according to the node fault matching degree. If R(i)<r1, the engine fault warning module performs a first-level warning for the engine fault. If r1≤R(i)<r2, the engine fault warning module performs a second-level warning for the engine fault. If R(i)≥r2, the engine fault warning module does not warn of the engine fault, wherein r1 represents a first matching degree threshold, and r2 represents a second matching degree threshold.

[0015] The beneficial effects of the present invention are as follows: through the real-time collection of engine operation information, engine operating condition information and nitrogen oxide concentration by the engine data acquisition module, multi-dimensional data of engine operation can be covered to achieve all-round fault diagnosis, ensure the timeliness of fault warning, and achieve comprehensive, accurate and real-time diagnosis and warning of engine faults, thereby improving the safety and reliability of engine operation, thereby improving the system's analysis efficiency of engine data and improving the accuracy of engine fault warning. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0017] Figure 1 This is a schematic diagram of the structure of the engine fault intelligent diagnosis system based on the Internet of Things in this embodiment.

[0018] Figure 2 Schematic diagram of the structure of the node data analysis module of this embodiment.

[0019] Figure 3 Schematic diagram of the structure of the data extraction and analysis module of this embodiment.

[0020] Figure 4 Schematic diagram of the structure of the sample matching analysis module in this embodiment. DETAILED DESCRIPTION

[0021] In order to more clearly illustrate the present invention, the present invention is further described below in conjunction with preferred embodiments and accompanying drawings. Similar components in the accompanying drawings are represented by the same reference numerals. It should be understood by those skilled in the art that the content specifically described below is illustrative rather than restrictive, and should not be used to limit the scope of protection of the present invention.

[0022] It should be noted that, although the terms first, second, third, etc. may be used to describe in the embodiments of the present application, these descriptions should not be limited to these terms. These terms are only used to distinguish the descriptions. For example, without departing from the scope of the embodiments of the present application, the first may also be referred to as the second, and similarly, the second may also be referred to as the first.

[0023] See also Figure 1 As shown, it is an engine fault intelligent diagnosis system based on the Internet of Things in this embodiment, including: The engine data acquisition module is used to collect engine operation information, engine operating condition information and nitrogen oxide concentration in real time. The engine operation information includes speed, cylinder temperature, oil pressure, vibration frequency, load current and cold start time. The speed is in rpm, the cylinder temperature is in °C, the oil pressure is in MPa, the vibration frequency is in kHz, the load current is in A, and the cold start time is the time to start the engine when the engine water temperature is lower than 40 degrees Celsius. The engine operating condition information includes idle speed, small load, medium load and high load. The engine operating condition information is data for distinguishing the throttle opening range. When the engine is running, if the accelerator pedal is in a completely relaxed state, the engine operating condition information is idle speed; if the throttle opening is within 25%, the engine operating condition information is light load; if the throttle opening is in the range of 25%-85%, the engine operating condition information is medium load; if the throttle opening is greater than 85%, the engine operating condition information is high load; the unit of the nitrogen oxide concentration is ppm; the engine operation information, engine operating condition information and nitrogen oxide concentration are collected by sensors installed on the engine.

[0024] Specifically, the present invention is applied to the cloud of the engine data management system, and is used to analyze whether there is an abnormality in the engine during the high-load operation of the engine and issue an early warning. The engine data acquisition module collects engine operation information (speed, cylinder temperature, oil pressure, vibration frequency, load current, cold start time) and nitrogen oxide concentration in real time to ensure the comprehensiveness of the data, and uses high-precision sensors (such as infrared thermocouples and vibration sensors) to obtain key parameters, providing a high-quality data basis for subsequent analysis and improving the accuracy of fault diagnosis.

[0025] Please continue reading Figure 1 As shown, the engine fault intelligent diagnosis system based on the Internet of Things also includes: The data node construction module is used to construct data nodes to obtain node information, wherein the node information includes node rotation speed, node cylinder temperature, node oil pressure, node vibration frequency, node load current and node nitrogen oxide concentration. The data node construction module is connected to the engine data acquisition module.

[0026] Specifically, the data node construction module in this embodiment sets a data node every preset analysis cycle, and numbers the data nodes in chronological order to obtain a node number, and sets the node number to i, i∈N + , and the average values ​​of the engine operation information and the nitrogen oxide concentration between the data node and the adjacent previous data node are respectively used as the node information.

[0027] Specifically, the node vibration frequency in this embodiment is a set of vibration frequencies between a data node and an adjacent previous data node.

[0028] Specifically, in this embodiment, the preset analysis cycle length is set to 10 seconds. In this embodiment, there is no specific limitation on the setting of the preset analysis cycle length, and technical personnel in this field can set it freely. For example, the preset analysis cycle length can also be set to 3 seconds, 5 seconds, and 15 seconds, etc.

[0029] Specifically, in this embodiment, the data node construction module constructs data nodes according to a preset period, discretizes continuous data for subsequent analysis, uses node numbers and timestamps to ensure data timing, provides clear timing features for fault diagnosis, and avoids data redundancy and noise interference.

[0030] Please continue reading Figure 1 As shown, the engine fault intelligent diagnosis system based on the Internet of Things also includes: The node data analysis module is used to analyze the main vibration frequency, node-related parameters, node pressure coefficient and node pressure theoretical value according to the node information, and to analyze the node data type according to the main vibration frequency, node-related parameters, node pressure coefficient and node pressure theoretical value. The node data type includes node vibration type, node temperature type and node pressure type. The node data analysis module is connected to the data node construction module.

[0031] See also Figure 2 As shown, the node data analysis module includes: The vibration main frequency analysis unit is used to analyze the vibration main frequency according to the node vibration frequency.

[0032] Specifically, the vibration main frequency analysis unit in this embodiment uses fast Fourier transform to extract the vibration main frequency from the node vibration frequency.

[0033] Please continue reading Figure 2 As shown, the node data analysis module also includes: The vibration type analysis unit is used to analyze the node vibration type according to the main vibration frequency, and the vibration type analysis unit is connected to the main vibration frequency analysis unit.

[0034] Specifically, the vibration type analysis unit in this embodiment analyzes the low-frequency amplitude according to the amplitude of the frequency point in the 0-200 Hz frequency band of the main vibration frequency, and sets the low-frequency amplitude to F low (i) Set The vibration type analysis unit analyzes the intermediate frequency amplitude according to the amplitude of the frequency point in the 200Hz-2kHz frequency band of the main vibration frequency, and sets the intermediate frequency amplitude to F mid (i) Set The vibration type analysis unit analyzes the high frequency amplitude according to the amplitude of the frequency point in the 2-5kHz frequency band of the main vibration frequency, and sets the high frequency amplitude to F high (i) Set , among which, A(i) k1 represents the amplitude of the frequency point in the 0-200Hz frequency band of the main vibration frequency, k1 represents the frequency point in the 0-200Hz frequency band of the main vibration frequency, Nk1(i) represents the number of frequency points in the 0-200Hz frequency band of the main vibration frequency, A(i) k2 represents the amplitude of the frequency point in the 200Hz-2kHz frequency band of the main vibration frequency, k2 represents the frequency point in the 200Hz-2kHz frequency band of the main vibration frequency, Nk2(i) represents the number of frequency points in the 200Hz-2kHz frequency band of the main vibration frequency, A(i) k3 represents the amplitude of the frequency point in the 2-5kHz frequency band of the main vibration frequency, k3 represents the frequency point in the 2-5kHz frequency band of the main vibration frequency, and Nk3(i) represents the number of frequency points in the 2-5kHz frequency band of the main vibration frequency.

[0035] Specifically, the vibration type analysis unit in this embodiment analyzes the node vibration type according to the low frequency amplitude, the medium frequency amplitude and the high frequency amplitude. If V(i) < v1 and F high (i) / F high (i-1)>fv, the vibration type analysis unit determines that the node vibration type is a type, if V(i)<v1 and F high (i) / F high (i-1)≤fv, the vibration type analysis unit determines that the node vibration type is Class II, if v1≤V(i)≤v2, the vibration type analysis unit determines that the node vibration type is Class II, if V(i)>v2, the vibration type analysis unit determines that the node vibration type is Class I, wherein V(i) represents the amplitude parameter, , v1 represents the first amplitude threshold, 1.1≤v1≤1.3, v2 represents the second amplitude threshold, 3.4≤v2≤3.6, fv represents the amplitude comparison threshold, 1.2≤fv≤1.3. It can be understood that the values ​​of the amplitude threshold and the amplitude comparison threshold are not specifically limited in this embodiment, and those skilled in the art can freely set them as long as they meet the analysis of the node vibration type. The optimal values ​​of the amplitude threshold and the amplitude comparison threshold are: v1=1.2, v2=3.5, fv=1.2.

[0036] Specifically, in this embodiment, the vibration type analysis unit analyzes the low-frequency, medium-frequency, and high-frequency amplitudes to determine the vibration type, thereby effectively analyzing the faults caused by the vibration of engine parts.

[0037] Please continue reading Figure 2 As shown, the node data analysis module also includes: The temperature and current analysis unit is used to analyze the node-related parameters according to the node cylinder temperature and the node load current, and the temperature and current analysis unit is connected to the vibration type analysis unit.

[0038] Specifically, the temperature current analysis unit in this embodiment analyzes the node-related parameters according to the node cylinder temperature and the node load current, sets the node-related parameters to Q(i), and sets Q(i)=[T(i)-T(i-1)]×e tl×T(i) / [t×L(i)], where T(i) represents the node cylinder temperature, L(i) represents the node load current, t represents the duration of the preset analysis cycle, tl represents the temperature parameter, and 0.01≤tl≤0.03. It can be understood that the value of the temperature parameter is not specifically limited in this embodiment, and those skilled in the art can freely set it as long as the analysis of the node-related parameters is satisfied. The optimal value of the temperature parameter is: tl=0.02.

[0039] Specifically, in this embodiment, the temperature-current analysis unit combines the in-cylinder temperature and the load current to calculate the node association parameters to quantify the correlation between temperature and load, provide a basis for temperature type analysis, and improve the accuracy of combustion state diagnosis.

[0040] Please continue reading Figure 2 As shown, the node data analysis module also includes: The temperature type analysis unit is used to analyze the node temperature type according to the node association parameter and the node nitrogen oxide concentration, and the temperature type analysis unit is connected to the temperature current analysis unit.

[0041] Specifically, the temperature type analysis unit described in this embodiment analyzes the node temperature type according to the node association parameters and the node nitrogen oxide concentration. If Q(i)<q1 and NO(i)>n, the temperature type analysis unit determines that the node temperature type is Class I; if Q(i)<q1 and NO(i)≤qn, the temperature type analysis unit determines that the node temperature type is Class II; if q1≤Q(i)≤q2, the temperature type analysis unit determines that the node temperature type is Class II; if Q(i)>q2, the temperature type analysis unit determines that the node temperature type is Class I, wherein q1 represents the first association threshold, 0.4≤q1≤0.6, q2 represents the second association threshold, 1.7≤q2≤1.9, and qn represents the nitrogen oxide threshold, 450≤qn≤550. It can be understood that the present embodiment does not impose any specific limitation on the values ​​of the association threshold and the nitrogen and oxygen threshold, and those skilled in the art can freely set them as long as they satisfy the analysis of the node temperature type. The optimal values ​​of the association threshold and the nitrogen and oxygen threshold are: q1=0.5, q2=1.8, qn=500.

[0042] Specifically, in this embodiment, the temperature type analysis unit analyzes the node-related parameters and the nitrogen oxide concentration to determine the temperature type, thereby realizing the fault analysis of the combustion anomaly and the cooling system fault.

[0043] Please continue reading Figure 2 As shown, the node data analysis module also includes: The pressure data analysis unit is used to analyze the node pressure coefficient and the node pressure theoretical value according to the node oil pressure, and the pressure data analysis unit is connected to the temperature type analysis unit.

[0044] Specifically, the pressure data analysis unit described in this embodiment uses the node number corresponding to the cold start time as the cold start node number, and analyzes the node pressure coefficient according to the cold start node number and the node oil pressure, sets the initial pressure to P0, sets P0=P(i=j), calculates the engine running time time(i), sets time(i)=(ij)×t, and makes the engine running time correspond to the node oil pressure one by one to obtain multiple sets of fitting data (time(i), P(i)), sets the attenuation curve to lnP(i)=lnP0-k×time(i), and substitutes the fitting data to obtain the node pressure coefficient k, wherein P(i) represents the node oil pressure and j represents the cold start node number.

[0045] Specifically, the calculation method of the node pressure coefficient k is not specifically limited in this embodiment, and those skilled in the art can freely set it, such as using the least squares method to solve k.

[0046] Specifically, the pressure data analysis unit in this embodiment analyzes the node pressure theoretical value according to the node oil pressure, the node speed and the node pressure coefficient, and sets the node pressure theoretical value to P1(i), and sets P1(i)=P0×e -k×time(i) +1.5×N(i)×[P(i)-P0] / {[N(i)-N(i=j)]+1}, where N(i) represents the node speed.

[0047] Specifically, in this embodiment, the pressure data analysis unit analyzes the cold start time and the oil pressure to analyze and calculate the node pressure coefficient and the node pressure theoretical value, thereby quantifying the health status of the oil system and analyzing faults such as oil pump wear and filter element blockage.

[0048] Please continue reading Figure 2 As shown, the node data analysis module also includes: The pressure type analysis unit is used to analyze the node pressure type according to the node pressure coefficient and the node pressure theoretical value, and the pressure type analysis unit is connected to the pressure data analysis unit.

[0049] Specifically, the pressure type analysis unit in this embodiment analyzes the node pressure type according to the node pressure coefficient and the node pressure theoretical value. If |P1(i) / P(i)-1|≤p and k>β, the pressure type analysis unit determines that the node pressure type is Class II. If |P1(i) / P(i)-1|>p and k>β, the pressure type analysis unit determines that the node pressure type is Class I. If k≤β, the pressure type analysis unit determines that the node pressure type is Class II, wherein p represents the pressure deviation threshold, 0.1≤p≤0.2, and β represents the pressure coefficient threshold, 0.001≤β≤0.002. It can be understood that the values ​​of the pressure deviation threshold and the pressure coefficient threshold are not specifically limited in this embodiment, and those skilled in the art can freely set them as long as they meet the analysis of the node pressure type. The optimal values ​​of the pressure deviation threshold and the pressure coefficient threshold are: p=0.15, β=0.0015.

[0050] Specifically, in this embodiment, the pressure type analysis unit analyzes the pressure coefficient and the theoretical value to determine the pressure type, thereby accurately analyzing the oil system failure and avoiding engine damage caused by abnormal oil pressure.

[0051] Please continue reading Figure 1 As shown, the engine fault intelligent diagnosis system based on the Internet of Things also includes: The data extraction and analysis module is used to extract node information according to the node data type and the engine operating condition information to obtain sample information, and analyze the sample node characteristics according to the sample information. The data extraction and analysis module is connected to the node data analysis module.

[0052] See also Figure 3 As shown, the data extraction and analysis module includes: The sample node extraction unit is used to extract the node information according to the node data type and the engine operating condition information to obtain the sample information.

[0053] Specifically, the sample node extraction unit in this embodiment extracts node information whose engine operating condition information is idle and whose node vibration type, node temperature type and node pressure type are all of Class II as idle sample information; extracts node information whose engine operating condition information is medium load and whose node vibration type, node temperature type and node pressure type are at least of Class I as first medium load sample information; extracts node information whose engine operating condition information is medium load and whose node vibration type, node temperature type and node pressure type are all of Class II as second medium load sample information; extracts node information whose engine operating condition information is high load and whose node vibration type, node temperature type and node pressure type are at least of Class I as first high load sample information; extracts node information whose engine operating condition information is high load and whose node vibration type, node temperature type and node pressure type are all of Class II as second high load sample information.

[0054] Specifically, in this embodiment, the sample node extraction unit analyzes the operating condition information and the node type to extract sample information, thereby ensuring that the sample data is representative and providing high-quality input for subsequent feature analysis.

[0055] Please continue reading Figure 3 As shown, the data extraction and analysis module also includes: The sample feature analysis unit is used to analyze the sample node features according to the sample information, and the sample feature analysis unit is connected to the sample node extraction unit.

[0056] Specifically, the sample feature analysis unit in this embodiment analyzes the sample node feature according to the sample information, sets the sample node feature as H(u), and sets , where u represents the sample data number, u∈{1,2,3,4,5}. When u=1, the idle sample information is used to analyze the sample node characteristics. H(1) represents the idle node characteristics. U(1) represents the set of node numbers corresponding to the idle sample information. NU(1) represents the number of idle sample information. When u=2, the first medium load sample information is used to analyze the sample node characteristics. H(2) represents the first medium load node characteristics. U(2) represents the set of node numbers corresponding to the first medium load sample information. NU(2) represents the number of first medium load sample information. When u=3, the second medium load sample information is used to analyze the sample node characteristics. H(3) represents the second medium load node characteristics. The sample node characteristics of the first high load are analyzed using the first high load sample information. U(3) represents the set of node numbers corresponding to the second medium load sample information. NU(3) represents the number of the second medium load sample information. When u=4, the sample node characteristics of the first high load are analyzed using the first high load sample information. H(4) represents the first high load node characteristics. U(4) represents the set of node numbers corresponding to the first high load sample information. NU(4) represents the number of the first high load sample information. When u=5, the sample node characteristics of the second high load are analyzed using the second high load sample information. H(5) represents the second high load node characteristics. U(5) represents the set of node numbers corresponding to the second high load sample information. NU(5) represents the number of the second high load sample information.

[0057] Please continue reading Figure 1 As shown, the engine fault intelligent diagnosis system based on the Internet of Things also includes: The similarity analysis module is used to analyze the load similarity according to the sample node characteristics, and the similarity analysis module is connected to the data extraction and analysis module.

[0058] Specifically, the similarity analysis module described in this embodiment analyzes the load similarity according to the sample node characteristics, sets the load similarity to S, and sets S=H(1)·H(3)·H(5) / {[H(1)+H(2)]·[H(1)+H(4)]}.

[0059] Specifically, in this embodiment, the load similarity analysis module is used to analyze the similarity of the engine states under different working conditions to quantify the similarity, provide a reference for fault matching analysis, and improve the robustness of fault diagnosis.

[0060] Please continue reading Figure 1 As shown, the engine fault intelligent diagnosis system based on the Internet of Things also includes: The sample matching analysis module is used to analyze the node fault matching degree according to the node information, node data type, engine operating condition information, load similarity, sample information and sample node characteristics. The sample matching analysis module is connected to the similarity analysis module.

[0061] See also Figure 4 As shown, the sample matching analysis module includes: The matching degree analysis unit is used to analyze the node fault matching degree according to the node information, engine operating condition information and sample node characteristics.

[0062] Specifically, the matching degree analysis unit in this embodiment analyzes the node fault matching degree according to the node information, the engine operating condition information and the sample node characteristics, and sets the fault matching degree to R(i). If the engine operating condition information is idle, set If the engine operating condition information is medium load or high load, set , where Max() is the maximum value extraction function, which is used to extract the maximum value among multiple groups of data in brackets.

[0063] Specifically, in this embodiment, the matching degree analysis unit analyzes the node information and sample node characteristics to analyze the fault matching degree, thereby quantifying the matching degree between the current state and the fault mode, and providing a basis for fault warning.

[0064] Please continue reading Figure 4 As shown, the sample matching analysis module also includes: The similarity association analysis unit is used to improve the analysis process of node fault matching according to load similarity to improve the diagnosis accuracy and avoid false positives and false negatives. The similarity association analysis unit is connected to the matching analysis unit.

[0065] Specifically, the similarity association analysis unit in this embodiment improves the analysis process of the node fault matching degree according to the load similarity. If S>s, the similarity association analysis unit does not improve the analysis process of the node fault matching degree; otherwise, the similarity association analysis unit improves the analysis process of the node fault matching degree, and the improved node fault matching degree is R1(i), and R1(i)=R(i)×e S ; Wherein, s represents the similarity threshold, 0.6≤s≤0.9. It can be understood that the value of the similarity threshold is not specifically limited in this embodiment, and those skilled in the art can freely set it, as long as the improvement of the analysis process of node fault matching is satisfied. The best value of the similarity threshold is: s=0.7.

[0066] Please continue reading Figure 4 As shown, the sample matching analysis module also includes: The type matching analysis unit is used to improve the analysis process of load similarity according to the node data type and sample information, and further optimize the accuracy of fault diagnosis. The type matching analysis unit is connected to the similarity association analysis unit.

[0067] Specifically, when at least one of the node vibration type, node temperature type and node pressure type in the current analysis cycle is of the same type, the type matching analysis unit in this embodiment counts the number of analysis cycles after the analysis cycle with node number j whose node data type is completely the same as the node data type of the current analysis cycle as the number of similar cycles, and improves the load similarity analysis process according to the number of similar cycles. The improved load similarity is S1, and S1=S×e NS(i) / (i-j+1) ; where NS(i) represents the number of similar cycles.

[0068] Please continue reading Figure 1 As shown, the engine fault intelligent diagnosis system based on the Internet of Things also includes: The engine fault warning module is used to warn of engine faults according to the node fault matching degree, and the engine fault warning module is connected to the sample matching analysis module.

[0069] Specifically, the engine fault warning module in this embodiment warns of engine faults according to the node fault matching degree. If R(i)<r1, the engine fault warning module warns of the engine fault at the first level. If r1≤R(i)<r2, the engine fault warning module warns of the engine fault at the second level. If R(i)≥r2, the engine fault warning module does not warn of the engine fault. Among them, r1 represents the first matching degree threshold, 0.2≤r1≤0.4, and r2 represents the second matching degree threshold, 0.5≤r2≤0.7. It can be understood that in this embodiment, the value of the matching degree threshold is not specifically limited, and those skilled in the art can freely set it as long as the warning of the engine fault is satisfied. The optimal value of the matching degree threshold is: r1=0.3, r2=0.6.

[0070] Specifically, when the engine fault warning module described in this embodiment warns of an engine fault, a first-level warning indicates that there is a serious fault in the engine, and the user should immediately stop using the engine and send it for inspection and repair. A second-level warning indicates that there may be a fault in the engine, and the user should check the engine to ensure that the engine is running well, or send it for inspection and repair while ensuring safety.

[0071] Specifically, in this embodiment, the engine fault warning module analyzes the fault matching degree to perform graded warnings, provide clear fault levels and response suggestions, and ensure that users take timely measures to avoid serious engine damage.

[0072] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not limitations on the embodiments of the present invention. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is impossible to list all the embodiments here. All obvious changes or modifications derived from the technical solution of the present invention are still within the protection scope of the present invention.

Claims

1. An intelligent engine fault diagnosis system based on the Internet of Things, characterized in that: include: Engine data acquisition module, used to collect engine operation information, engine operating condition information and nitrogen oxide concentration in real time; Data node construction module, used to construct data nodes to obtain node information; A data extraction and analysis module is used to extract node information according to the node data type and the engine operating condition information to obtain sample information, and analyze the sample node characteristics according to the sample information; the data extraction and analysis module includes a sample node extraction unit and a sample characteristic analysis unit, the sample node extraction unit is used to extract node information according to the node data type and the engine operating condition information to obtain sample information, and the sample characteristic analysis unit is used to analyze the sample node characteristics according to the sample information; A sample matching analysis module is used to analyze the node fault matching degree according to the node information, node data type, engine operating condition information, load similarity, sample information and sample node characteristics; The engine fault warning module is used to warn of engine faults according to the node fault matching degree.

2. The engine fault intelligent diagnosis system based on the Internet of Things according to claim 1 is characterized in that: The system also includes: a node data analysis module, which is used to analyze the main vibration frequency, node-related parameters, node pressure coefficient and node pressure theoretical value according to the node information, and analyze the node data type according to the main vibration frequency, node-related parameters, node pressure coefficient and node pressure theoretical value. The node data analysis module includes a vibration main frequency analysis unit, a vibration type analysis unit, a temperature current analysis unit, a temperature type analysis unit, a pressure data analysis unit and a pressure type analysis unit.

3. The engine fault intelligent diagnosis system based on the Internet of Things according to claim 2 is characterized in that: The vibration main frequency analysis unit extracts the vibration main frequency from the node vibration frequency using fast Fourier transform; The vibration type analysis unit analyzes the low frequency amplitude, the medium frequency amplitude and the high frequency amplitude according to the amplitude of the frequency points in different frequency bands in the main vibration frequency, and determines the node vibration type according to the low frequency amplitude, the medium frequency amplitude and the high frequency amplitude, and the node vibration type includes one category and two categories; The temperature and current analysis unit analyzes the node-related parameters according to the node cylinder temperature and the node load current to obtain the node-related parameters Q(i); The temperature type analysis unit determines the node temperature type according to the node association parameter and the node nitrogen oxide concentration, and the node temperature type includes type I and type II; The pressure data analysis unit uses the node number corresponding to the cold start time as the cold start node number, and analyzes the node pressure coefficient according to the cold start node number and the node oil pressure to obtain the node pressure coefficient k; The pressure data analysis unit analyzes the node pressure theoretical value according to the node oil pressure, the node speed and the node pressure coefficient to obtain the node pressure theoretical value P1(i); The pressure type analysis unit determines the node pressure type according to the node pressure coefficient and the node pressure theoretical value, and the node pressure type includes one type and two types.

4. The engine fault intelligent diagnosis system based on the Internet of Things according to claim 3 is characterized in that: The sample node extraction unit extracts node information whose engine operating condition information is idle and whose node vibration type, node temperature type and node pressure type are all of the second category as idle sample information; extracts node information whose engine operating condition information is medium load and whose node vibration type, node temperature type and node pressure type are at least of the first category as first medium load sample information; extracts node information whose engine operating condition information is medium load and whose node vibration type, node temperature type and node pressure type are all of the second category as second medium load sample information; extracts node information whose engine operating condition information is high load and whose node vibration type, node temperature type and node pressure type are at least of the first category as first high load sample information; extracts node information whose engine operating condition information is high load and whose node vibration type, node temperature type and node pressure type are all of the second category as second high load sample information.

5. The engine fault intelligent diagnosis system based on the Internet of Things according to claim 4 is characterized in that: The sample feature analysis unit analyzes the sample node feature according to the sample information to obtain the sample node feature H(u), where u represents the sample data number, u∈{1,2,3,4,5}.

6. The engine fault intelligent diagnosis system based on the Internet of Things according to claim 5 is characterized in that: The system further comprises: a similarity analysis module for analyzing load similarity according to sample node characteristics; The similarity analysis module analyzes the load similarity according to the sample node feature H(u) to obtain the load similarity S.

7. The engine fault intelligent diagnosis system based on the Internet of Things according to claim 6 is characterized in that: The sample matching analysis module is provided with a matching degree analysis unit, which is used to analyze the node fault matching degree under different engine operating condition information according to the node information and the sample node feature H(u) to obtain the fault matching degree R(i).

8. The engine fault intelligent diagnosis system based on the Internet of Things according to claim 7 is characterized in that: The sample matching analysis module is further provided with a similarity association analysis unit, which is used to improve the analysis process of node fault matching according to load similarity, and the improved node fault matching is R1(i).

9. The engine fault intelligent diagnosis system based on the Internet of Things according to claim 8 is characterized in that: The sample matching analysis module is also provided with a type matching analysis unit, which is used to count the number of analysis cycles whose node data types are exactly the same as the node data types of the current analysis cycle after the analysis cycle with node number j as the number of similar cycles when at least one of the node vibration type, node temperature type and node pressure type of the current analysis cycle is of the same type, and improve the analysis process of load similarity according to the number of similar cycles, and the improved load similarity is S1.

10. The engine fault intelligent diagnosis system based on the Internet of Things according to claim 9 is characterized in that: The engine fault warning module warns of engine fault according to the node fault matching degree. If R(i)<r1, the engine fault warning module performs a first-level warning for the engine fault. If r1≤R(i)<r2, the engine fault warning module performs a second-level warning for the engine fault. If R(i)≥r2, the engine fault warning module does not warn of the engine fault, wherein r1 represents a first matching degree threshold, and r2 represents a second matching degree threshold.

Images