Traction fan fault diagnosis test bench based on multi-source signal fusion

Abstract

The invention discloses a traction fan fault diagnosis test bed based on multi-source signal fusion. The test bed comprises a case for simulating a real traction system, a fan assembly, a data acquisition assembly comprising various sensors, a simulation assembly and a processing assembly. Partitions are arranged in the machine case, the real structure and the heat dissipation effect of the traction system machine case are simulated, traction fan operation data closer to the real working environment can be collected, meanwhile, multiple different types of sensors are arranged to collect multi-source data, the collected data are fused through processing equipment, and the data collection efficiency is improved. According to the invention, a plurality of signal fault features such as sound, current, vibration and the like are learned in cooperation with algorithms such as a deep learning network, so that an autonomous diagnosis effect can be formed, and compared with the prior art, signal acquisition comprehensiveness and automation of signal transmission, processing and analysis in a traction fan fault research process can be improved; therefore, researchers can be assisted to improve the efficiency and accuracy of fault diagnosis and research of the traction fan.

Description

Technical Field

[0001] The present invention relates to the technical field of fan fault diagnosis, and in particular to a traction fan fault diagnosis test bench based on multi-source signal fusion. Background Art

[0002] The traction system of the EMU is responsible for providing driving force for the entire EMU. The traction system accumulates a large amount of heat during operation, and cooling fans and other heat dissipation devices are required to dissipate the heat in time to ensure the normal operation of the traction system. Therefore, how to ensure the normal operation of the cooling fan and eliminate the fault of the cooling fan in time is an important part of maintaining the continuous operation of the traction system.

[0003] It is difficult for ordinary companies to obtain relevant data on the factors of cooling fan failure. The conventional method is to collect data on the exposed fan, but its authenticity is divorced from the actual working conditions. Although the establishment of a complete traction system can obtain simulated data, the cost is very expensive. In addition, in the above data acquisition process, there are still limitations such as the single type of data signal collection and the backward network model. Therefore, it is of great significance to simplify the internal structure and required components of the traction system while ensuring the working effect of the cooling system, develop an economically feasible fan test bench, design a reasonable test plan, and collect the cooling fan operation data closest to the real working environment.

[0004] In the prior art, the fault diagnosis research of traction fans is carried out by collecting the vibration data of the fan bearings for analysis, and then transferring the recorded data from the collection equipment to the computing workstation for algorithm research. The operation is complicated and inefficient, the collected data signals are of a single type and the data cannot simulate the actual situation, which cannot meet the research needs of fault diagnosis.

[0005] Therefore, the prior art still needs to be improved and developed. Summary of the invention

[0006] In order to solve the problems in the prior art of fault diagnosis research on traction fans, the vibration data of fan bearings is collected for analysis, and then the recorded data is transferred from the collection device to the computing workstation for algorithm research. The operation is complicated and inefficient, the collected data signals are of a single type and the data cannot simulate the actual situation, and the research needs of fault diagnosis cannot be met. The present invention proposes a traction fan fault diagnosis test bench based on multi-source signal fusion.

[0007] The present invention is achieved through the following technical solutions:

[0008] A traction fan fault diagnosis test bench based on multi-source signal fusion, wherein the traction fan fault diagnosis test bench based on multi-source signal fusion comprises:

[0009] A chassis, wherein a component chamber and a fan chamber are isolatedly arranged inside the chassis, and the fan chamber is connected to the component chamber;

[0010] A fan assembly, the fan assembly is embedded in the fan chamber;

[0011] A data acquisition component, the data acquisition component comprising a vibration sensor, a sound sensor arranged on the fan component, a temperature sensor and a current sensor arranged in the component room;

[0012] A simulation component, the simulation component comprising a heating mechanism disposed in the component chamber;

[0013] A processing component, wherein the processing component comprises a PLC module, and the processing component is arranged in the component room and is connected to the fan component, the data acquisition component and the simulation component through a circuit.

[0014] The traction fan fault diagnosis test bench based on multi-source signal fusion, wherein the component chamber comprises a first component chamber and a second component chamber which are arranged on both sides of the fan chamber and communicated with the fan chamber;

[0015] A plurality of first ventilation holes are arranged on the side wall of the chassis corresponding to the first component chamber;

[0016] A plurality of second ventilation holes are arranged on the side wall of the chassis corresponding to the second element chamber.

[0017] The traction fan fault diagnosis test bench based on multi-source signal fusion, wherein the fan chamber is arranged through two opposite side walls of the chassis;

[0018] The chassis is detachably provided with a fan chamber baffle, and the fan chamber baffle covers both ends of the fan chamber.

[0019] The traction fan fault diagnosis test bench based on multi-source signal fusion, wherein a first air guide port is provided at the bottom of the fan chamber, and a space corresponding to the first air guide port is connected to the first component chamber;

[0020] The first element chamber is provided with an air duct top plate connected to the side wall of the chassis;

[0021] A plurality of air duct ribs connected to the air duct top plate and the chassis bottom plate are arranged below the air duct top plate, and hollowed-out guide holes are arranged on the plurality of air duct ribs.

[0022] The traction fan fault diagnosis test bench based on multi-source signal fusion, wherein a filter is provided in the first element chamber at a position corresponding to the air duct rib;

[0023] A plurality of second air guide ports are arranged on the side wall of the second element chamber corresponding to the fan chamber, and a dustproof net is embedded in the second air guide port.

[0024] The traction fan fault diagnosis test bench based on multi-source signal fusion, wherein the fan component includes:

[0025] An outer rotor centrifugal fan, the vibration sensor and the sound sensor are fixedly arranged on the outer rotor centrifugal fan;

[0026] A frame, wherein the frame is connected to the bottom plate of the fan chamber by bolts, the outer rotor centrifugal fan is fixedly connected to the frame, and the outer rotor centrifugal fan corresponds to the position of the first air guide port.

[0027] The traction fan fault diagnosis test bench based on multi-source signal fusion, wherein the heating mechanism is provided with two, and the two heating mechanisms are respectively fixedly arranged in the first element chamber and the second element chamber;

[0028] Two temperature sensors are provided, and the two temperature sensors are fixedly disposed in the first element chamber and the second element chamber respectively.

[0029] The traction fan fault diagnosis test bench based on multi-source signal fusion, wherein the processing component comprises:

[0030] The PLC module is arranged in the first element chamber;

[0031] An edge calculator, the edge calculator is arranged in the second element chamber, and the edge calculator is connected to the PLC module through a circuit;

[0032] The data acquisition component also includes a data collector, in which the current sensor is integrated, the vibration sensor, the sound sensor and the temperature sensor are connected to the data collector via a circuit, and the data collector is connected to the edge calculator via a circuit.

[0033] The traction fan fault diagnosis test bench based on multi-source signal fusion is characterized in that a touch display is also embedded on the side wall of the chassis corresponding to the second element chamber, and the touch display is connected to the edge calculator through a circuit, and the touch display is used to feedback data information.

[0034] The traction fan fault diagnosis test bench based on multi-source signal fusion, wherein the opposite ends of the chassis corresponding to the first component chamber and the second component chamber are open, and the chassis includes two chassis side panels, and the two chassis side panels are respectively detachably connected to the open parts at the opposite ends of the chassis.

[0035] The beneficial effects of the present invention are as follows: the present invention can collect traction fan operation data that is closer to the actual working environment by arranging partitions inside the chassis to simulate the actual structure and heat dissipation effect of the traction system chassis. At the same time, a variety of different types of sensors are arranged to collect multi-source data. The collected data is fused with multiple sources through processing equipment, and the software such as deep learning networks is used to learn the fault characteristics of various signals such as sound, current, and vibration, thereby forming an autonomous diagnosis effect. Compared with the existing technology, the comprehensiveness of signal collection in the process of traction fan fault research, the automation of signal transmission, processing and analysis, can be improved, thereby assisting researchers to improve the efficiency and accuracy of traction fan fault diagnosis research. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 It is a structural explosion diagram of the traction fan fault diagnosis test bench based on multi-source signal fusion of the present invention;

[0037] Figure 2 It is a structural exploded top view of the traction fan fault diagnosis test bench based on multi-source signal fusion of the present invention;

[0038] Figure 3 It is a top view of the installation state of the traction fan fault diagnosis test bench based on multi-source signal fusion of the present invention;

[0039] Figure 4 It is a chassis structure diagram of a traction fan fault diagnosis test bench based on multi-source signal fusion of the present invention;

[0040] Figure 5 It is a schematic structural diagram of a fan component of a traction fan fault diagnosis test bench based on multi-source signal fusion of the present invention;

[0041] Figure 6 It is a hardware structure diagram of a PLC module of a traction fan fault diagnosis test bench based on multi-source signal fusion according to the present invention;

[0042] Figure 7 It is a schematic diagram of a heating mechanism of a traction fan fault diagnosis test bench based on multi-source signal fusion of the present invention;

[0043] Figure 8 It is a schematic diagram of the structure of the temperature sensor of the traction fan fault diagnosis test bench based on multi-source signal fusion of the present invention;

[0044] Fig. 9 It is a schematic diagram of the structure of the sound sensor of the traction fan fault diagnosis test bench based on multi-source signal fusion of the present invention;

[0045] exist Figures 1 to 9In: 100, chassis; 101, fan room baffle; 102, chassis side panel; 110, component room; 111, first component room; 111a, first ventilation hole; 112, second component room; 113, air duct top plate; 114, air duct ribs; 115, guide hole; 120, fan room; 121, first air guide port; 130, filter; 140, dust net; 200, fan assembly; 210, outer rotor centrifugal fan; 220, rack; 300, data acquisition assembly; 310, vibration sensor; 320, sound sensor; 330, temperature sensor; 340, data collector; 400, analog assembly; 410, heating mechanism; 500, processing assembly; 510, edge calculator; 520, PLC module; 600, touch display. DETAILED DESCRIPTION

[0046] In order to make the purpose, technical solution and effect of the present invention clearer and more specific, the present invention is further described in detail with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0047] It should be noted that if the embodiments of the present invention involve directional indications (such as up, down, left, right, front, back, etc.), the directional indications are only used to explain the relative position relationship, movement status, etc. between the components under a certain specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indications will also change accordingly.

[0048] In addition, if there are descriptions involving "first", "second", etc. in the embodiments of the present invention, the descriptions of "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or suggesting their relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In addition, the technical solutions between the various embodiments can be combined with each other, but they must be based on the ability of ordinary technicians in the field to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0049] In the prior art, the fault diagnosis research of traction fans is carried out by collecting the vibration data of the fan bearings for analysis, and then transferring the recorded data from the collection device to the computing workstation for algorithm research. The operation is complicated and inefficient, the collected data signals are of a single type and the collection conditions are divorced from the actual situation, which cannot meet the research needs of fault diagnosis.

[0050] Based on the above problems in the prior art, the present invention provides a traction fan fault diagnosis test bench based on multi-source signal fusion, such as Figure 1As shown, the traction fan fault diagnosis test bench based on multi-source signal fusion includes: a chassis 100, wherein a component chamber 110 and a fan chamber 120 are isolated and arranged inside the chassis 100, and the fan chamber 120 is connected to the component chamber 110; a fan assembly 200, wherein the fan assembly 200 is embedded in the fan chamber 120; a data acquisition assembly 300, wherein the data acquisition assembly 300 includes a vibration sensor 310, a sound sensor 320 arranged on the fan assembly 200, and a temperature sensor 330 and a current sensor arranged in the component chamber 110; an analog assembly 400, wherein the analog assembly 400 includes a heating mechanism 410 arranged in the component chamber 110; and a processing assembly 500, wherein the processing assembly 500 includes a PLC module 520, and the processing assembly 500 is arranged in the component chamber 110 and is circuit-connected with the fan assembly 200, the data acquisition assembly 300 and the analog assembly 400.

[0051] The present invention simulates the actual structure and heat dissipation effect of the traction system chassis 100 by setting partitions inside the chassis 100, and can collect traction fan operation data that is closer to the actual working environment. At the same time, a variety of different types of sensors are set to collect multi-source data. The collected data is formed into multi-source fusion through processing equipment, and cooperates with software such as deep learning networks to learn various signal fault characteristics such as sound, current, and vibration, thereby forming an autonomous diagnosis effect. Compared with the existing technology, it can improve the comprehensiveness of signal collection in the process of traction fan fault research, and the automation of signal transmission, processing and analysis, thereby assisting researchers to improve the efficiency and accuracy of traction fan fault diagnosis research.

[0052] In the above embodiment, if Figure 1 and Figure 2 As shown, the main body of the traction fan fault diagnosis test bench based on multi-source signal fusion of the present invention is composed of a chassis 100, a fan assembly 200, a data acquisition assembly 300, a simulation assembly 400 and a processing assembly 500, wherein the structure of the chassis 100 simulates the structural relationship between the traction fan and the traction system, and a component chamber 110 and a fan chamber 120 are isolated and arranged inside the chassis 100, wherein the component chamber 110 is used to install functional components and is used to simulate the heating condition of the traction system, and the fan chamber 120 is used to install the fan assembly 200, thereby simulating the actual working state of the traction fan. In this embodiment, the fan chamber 120 is connected to the component chamber 110 to form a heat dissipation air duct path that simulates the traction assembly in the real EMU.

[0053] The fan assembly 200 is embedded in the fan chamber 120. The fan assembly 200 adopts a fan structure similar to that of a traction fan, which will be described in detail below.

[0054] The data acquisition component 300 is used to collect various data during the operation of the fan component 200. In the present embodiment, the data acquisition component 300 includes a vibration sensor 310, a sound sensor 320, a temperature sensor 330 and a current sensor arranged on the fan component 200, wherein the vibration sensor 310 is used to obtain the vibration condition of the fan component 200 to determine whether the fan has a fault; the sound sensor 320 is used to determine the sound changes when the fan blades are in the working state, thereby determining whether the fan has a fault by sound; the temperature sensor 330 is used to record the temperature changes of the component chamber 110, thereby determining the actual efficiency changes of the fan cooling by temperature monitoring; the current sensor is used to monitor the current supply condition of the fan component 200 to determine whether the fan component 200 is operating at the rated condition.

[0055] like Figure 3 and Figure 7 As shown, the simulation component 400 specifically includes a heating mechanism 410 arranged in the component chamber 110. The heating mechanism 410 generates heat to simulate the heat generated during the actual operation of the traction system. When the fan component 200 is in a working state, an air path can be formed through the chassis 100 to cool the component chamber 110. Due to the setting of the temperature sensor 330, it is convenient to monitor the changes in the cooling efficiency under different fan component states, thereby facilitating the study of the impact of fan failures on the changes in cooling efficiency under various conditions.

[0056] The processing component 500 is arranged in the component chamber 110 and is connected to the fan component 200, the data acquisition component 300 and the simulation component 400. The processing component 500 is used to control the working conditions of the fan component 200 and the simulation component 400, so as to simulate the heat generated by the real motor vehicle traction system and the cooling effect of the real traction fan. At the same time, by processing the data collected by various sensors in the data acquisition component 300, the data changes can be summarized and summarized. In this embodiment, the processing component 500 includes a PLC module 520. The PLC is a programmable logic controller (PLC), which is a digital operation electronic system specially designed for application in an industrial environment. It uses a programmable memory to store instructions for performing operations such as logic operations, sequential control, timing, counting and arithmetic operations, and controls various types of mechanical equipment or production processes through digital or analog input and output.

[0057] In this embodiment, by embedding a multi-source signal fusion algorithm and a deep learning network model in other functional elements of the PLC module 520 and the processing component 500, the signals obtained by the above-mentioned multiple sensors can be combined to form a fusion fault feature, thereby achieving full automation of multi-source signal data collection, transmission, processing, and analysis, improving fault diagnosis effects, and further assisting researchers in improving research efficiency and accuracy.

[0058] Further, in a specific embodiment, Figure 3 and Figure 4 As shown, there are two component chambers 110, namely a first component chamber 111 and a second component chamber 112. The first component chamber 111 and the second component chamber 112 are respectively arranged on both sides of the fan chamber 120 and are connected to the fan chamber 120. In this embodiment, the adjacent EMU traction system can be simulated by the first component chamber 111 and the second component chamber 112, and the influence range of the traction fan can be simulated by the fan assembly 200 in the fan chamber 120. In addition, since the first component chamber 111 and the second component chamber 112 are respectively connected to the fan chamber 120, an air duct that specifically simulates the heat dissipation process can be formed, thereby realizing the study of the influence of the fan on the airflow change.

[0059] In the specific configuration, since the chassis 100 is completely closed, in order to achieve air circulation, Figure 4 As shown, a plurality of first ventilation holes 111a are provided on the side wall of the first element chamber 111, and the plurality of first ventilation holes 111a are hollowed out and arranged in an array, and can be used for air intake. A plurality of second ventilation holes (not shown in the figure) are also provided on the side wall of the second element chamber 112, and the plurality of second ventilation holes are hollowed out and arranged in an array, and can be used for air exhaust, thereby forming a heat dissipation air duct connecting the inside of the chassis 100 with the outside of the chassis 100.

[0060] In another embodiment of the present invention, Figure 1 As shown, when the chassis 100 is specifically set up, in order to facilitate researchers to control the operating variables of the fan assembly 200 and test various factors that affect the operation of the fan, in this embodiment, the fan chamber 120 is set through the two opposite side walls of the chassis 100, so that a through space is formed in the middle of the chassis 100, and researchers can manually adjust the state of the fan in advance before the test, or apply an agent to affect the operating state of the fan assembly 200. At the same time, Figure 2 To prevent the fan assembly 200 in the fan chamber 120 from being affected by the external environment during operation, in this embodiment, a fan chamber baffle 101 is detachably provided on the chassis 100, and the fan chamber baffle 101 covers both ends of the fan chamber 120, so that the fan chamber 120 can be closed during the test, and the operation is convenient when adjusting the state of the fan assembly 200.

[0061] In another embodiment of the present invention, Figure 1 and Figure 4 As shown, a first air guide port 121 is provided at the bottom of the fan chamber 120, and a space is preset below the first air guide port 121, which is connected to the first element chamber 111. Correspondingly, an air duct top plate 113 connected to the side wall of the chassis 100 is provided inside the first element chamber 111, and the height of the air duct top plate 113 is located at the same horizontal plane as the plate at the bottom of the fan chamber 120. A space connected to the air duct top plate 113 and the bottom plate of the chassis 100 is also provided below the air duct top plate 113. The plurality of duct ribs 114 are used to support the duct top plate 113 to prevent it from being deformed due to the pressure of the components in the right component chamber 110. At the same time, the plurality of duct ribs 114 divide the bottom of the chassis 100 into a plurality of ventilation structures to together form an air duct. In order to prevent the duct ribs 114 from obstructing the air flow, in the present embodiment, guide holes 115 may be further provided on the duct ribs 114. The guide holes 115 are hollowed out to form a connection between the spaces between the duct ribs 114.

[0062] In this embodiment, the air duct top plate 113 is also used to install the temperature sensor 330 in the data acquisition component 300, the heating mechanism 410 in the simulation component 400, and the PLC module 520 in the processing component 500, etc.

[0063] Furthermore, if Figure 1 and Figure 3 As shown, a filter 130 is further provided at a position corresponding to the air duct rib 114 inside the above-mentioned first element chamber 111. When the fan assembly 200 starts to work, the fan assembly 200 generates suction negative pressure on the air duct during operation. The air enters the air duct through the first ventilation hole 111a of the first element chamber 111 through the first element chamber 111, and enters the fan chamber 120 after being filtered by the filter 130. In this process, the filter 130 plays the role of filtering dust and foreign matter to prevent dust and foreign matter from affecting the test results.

[0064] A plurality of second air ducts are provided on the side walls of the second element chamber 112 corresponding to the fan chamber 120, and a dustproof net 140 is embedded in the second air duct. When the fan assembly 200 starts to work, the air sucked into the fan chamber 120 passes through the dustproof net 140 of the second air duct and enters the second element chamber 112, and then is discharged through the second ventilation holes corresponding to the second element chamber 112. In this process, the dustproof net 140 further realizes the dustproof effect to prevent dust and the like from entering the second element chamber 112 and causing problems such as short circuit of the processing assembly 500 and affecting the sensitivity of the sensor.

[0065] In another embodiment of the present invention, Figure 5As shown, the fan assembly 200 specifically includes an outer rotor centrifugal fan 210 and a frame 220, wherein the frame 220 is fixedly connected to the bottom plate of the fan chamber 120 by bolts, and the outer rotor centrifugal fan is fixedly connected to the frame 220. During assembly, it should be ensured that the outer rotor centrifugal fan is in contact with the first air guide port 121 (such as Figure 1 As shown in the figure, the position of the air flow in the air duct can be driven with the highest efficiency when negative pressure is formed.

[0066] During installation, if Figure 1 and Fig. 9 As shown, the vibration sensor 310 and the sound sensor 320 are fixedly arranged on the outer rotor centrifugal fan. For the operation process of the outer rotor centrifugal fan, its dynamic balance level significantly affects the noise and life of the fan. When the dynamic balance is poor, the fan is at risk of causing resonance. The vibration sensor 310 can monitor the shaft rotation changes of the outer rotor centrifugal fan, so as to verify the dynamic balance state of the fan under different external conditions, so as to assist in studying the causes of the resonance and provide data for fault diagnosis. When the outer rotor centrifugal fan has problems in operation, the edges of its blades are prone to float up and down, and the fan shaft produces shaft noise. Such sounds are collected by the sound sensor 320, and combined with the corresponding data obtained by the vibration sensor 310, it is convenient to form a fault model for fan faults, so as to facilitate the study of the impact of the external environment and the corresponding state of the fan.

[0067] In another embodiment of the present invention, Figure 3 and Figure 8 As shown, two heating mechanisms 410 are provided, and the two heating mechanisms 410 are respectively fixedly arranged in the first element chamber 111 and the second element chamber 112, so as to simulate the heating conditions of the working systems at both ends of the traction fan. Correspondingly, two temperature sensors 330 are also provided, and the two temperature sensors 330 are respectively fixedly arranged in the first element chamber 111 and the second element chamber 112, so as to monitor the air temperature of the first element chamber 111 and the second element chamber 112, and obtain relevant data on the actual operating effect of the fan assembly.

[0068] Furthermore, the processing component 500 is specifically composed of the PLC module 520 and the edge calculator 510, wherein Figure 1 and Figure 6As shown, the PLC module 520 is arranged in the first element chamber 111. On the opposite side walls of the first element chamber 111 and the second element chamber 112, a clearance hole can be opened to facilitate connecting the PLC module 520 with the edge calculator 510 through wires. The working process of the PLC module 520 is generally divided into three stages, namely input sampling, user program execution and output refresh. Completing the above three stages is called a scan cycle. During the entire operation, the CPU of the programmable logic controller repeatedly executes the above three stages at a certain scan speed.

[0069] like Figure 1 As shown, the edge calculator 510 is arranged in the second element room 112, and the edge calculator 510 is circuit-connected to the PLC module 520. Edge computing refers to the use of an open platform integrating network, computing, storage, and application core capabilities to provide the nearest service on the side close to the object or data source. Its application is initiated on the edge side, resulting in a faster network service response, meeting the basic needs of the industry in terms of real-time business, application intelligence, security and privacy protection. The edge calculator 510 is pre-installed with a multi-source signal fusion algorithm and a deep learning network. The data obtained by the sensor is transmitted to the embedded data processing algorithm for signal fusion, and the processed data is input into the deep learning network model for training. The weights generated by the training can be used for prediction in the next step, so as to form a data correspondence relationship between various causes of traction fan failure.

[0070] Through the cooperation between the edge calculator 510 and the PLC module 520, researchers can obtain test results and various parameters at the fastest speed, avoiding the process of data migration and reprocessing in traditional technologies, and effectively improving the efficiency of traction fan fault diagnosis research.

[0071] In this embodiment, the data acquisition component 300 further includes a data collector 340, which is integrated with a data terminal using RFID technology and is a terminal computer device with a battery that can be operated offline. It has the functions of real-time acquisition, automatic storage, instant display, instant feedback, automatic processing, and automatic transmission. The data collector 340 is integrated with a current sensor, which can be used to monitor the current supply changes during the operation of the fan assembly, so as to correspond to the actual working state of the fan assembly. The vibration sensor 310, the sound sensor 320, and the temperature sensor 330 are connected to the data collector 340 circuit, and the data is summarized through the data collector 340 for the edge calculator 510 and the PLC module 520 to call.

[0072] In another embodiment of the present invention, Figure 1As shown, in order to facilitate researchers to quickly operate and intuitively obtain test data, a touch display 600 is also embedded on the side wall of the above-mentioned chassis 100 corresponding to the second element chamber 112. The touch display 600 is circuit-connected to the edge calculator 510. In actual use, the edge calculator 510 outputs information such as test data and results of traction fan fault diagnosis to the touch display 600. Researchers can directly obtain the test results and verify them through the synchronously output sensor data. Through the touch display 600, instructions can also be issued to the edge calculator 510 to control the working modes of the fan component 200 and the simulation component 400, so that rapid adjustments can be made to form the effect of controlling variable tests, effectively improving test efficiency.

[0073] In another embodiment of the present invention, Figure 1 As shown, the two opposite ends of the chassis 100 corresponding to the first component chamber 111 and the second component chamber 112 are open. At the same time, the chassis 100 also includes two chassis side panels 102, which are detachably connected to the open parts at the opposite ends of the chassis 100, so as to achieve the effect of rapid installation and disassembly. In actual use, it is convenient for researchers to replace or repair the functional components in the first component chamber 111 and the second component chamber 112.

[0074] In this embodiment, handles may be provided on the outer sides of the two chassis side panels 102 and the two fan chamber baffles 101 to facilitate researchers to disassemble or move the entire multi-source signal fusion traction fan fault diagnosis test bench.

[0075] Based on the above embodiments, the actual use process of the traction fan fault diagnosis test bench based on multi-source signal fusion of the present invention is as follows:

[0076] Sticking tape on the fan blades of the outer rotor centrifugal fan 210 to apply load to simulate dust accumulation / or using tape to cover part of the first ventilation holes 111a to simulate blockage;

[0077] Specifically, the test type is determined first. Common traction fan failures in engineering are blade uniform load, single blade eccentric load, and air inlet blockage. The dust accumulation on a single blade is usually less than 10g. Accordingly, the test types are determined as blade uniform load, single blade load, and air inlet blockage. The proposed weight of blade uniform load is six groups of 0g / 2g / 4g / 6g / 8g / 10g, the proposed weight of single blade eccentric load is four groups of 0g / 3g / 6g / 10g, and the proposed degree of air inlet blockage is five groups of 0 / 20% / 40% / 60% / 80%.

[0078] Then prepare the load. In order to reliably and effectively apply the load to the blade, the test uses tape to simulate dust accumulation, and the tape needs to be weighed. In the blade uniform load test, the tape needs to be attached to each blade to apply the load, and in the single blade eccentric load test, the tape is only attached to one blade to apply the load.

[0079] For the blocking test of the first ventilation holes 111a, it is only necessary to use tape to close a certain proportion of the air inlet holes.

[0080] Then, parameters such as the target temperature of the heating mechanism 410 in the first element chamber 111 and the second element chamber 112 and the rotation speed of the fan assembly 200 are set in the software panel of the touch screen;

[0081] Click Start in the software panel, the heating mechanism 410 and the temperature sensor 330 start working, and after the temperature in the component chamber 110 reaches the target temperature, the fan assembly 200 starts to dissipate heat;

[0082] Specifically, if the target temperature is set to 60°C and the speed of the fan assembly 200 is set to 3000 rpm, the heating mechanism 410 will stop working when the chamber temperature is heated to 60°C to avoid safety hazards or damage to components caused by excessively high temperatures in the chamber.

[0083] Then the fan assembly 200 starts to run at a constant speed of 3000 rpm. After the temperature drops below the target temperature, the heating mechanism 410 will start working again. Finally, the temperature in the first element chamber 111 and the second element chamber 112 will reach a dynamic equilibrium temperature lower than the target temperature. The dynamic equilibrium temperature is determined by the set rotation speed of the fan assembly 200.

[0084] The touch display 600 can be used to see the real-time temperature in the first element chamber 111 and the second element chamber 112 to observe whether the heat dissipation effect meets the requirements, so as to adjust the parameters to make them close to the actual working state of the fan assembly 200.

[0085] Click on the touch display 600 to collect data, and the data collector 340 controls the current, vibration, and sound sensors 320 to collect data;

[0086] Specifically, the edge computing device sends a collection instruction to the data collector 340, and the data collector 340 synchronously collects the three signal data of vibration, current, and sound during the operation of the fan assembly 200 through the vibration, current, and sound sensors 320. The collected data is stored in the training area or prediction area in the edge computing device, and the user can rename, export, delete, and perform other operations on it.

[0087] If there are multiple groups of tests, the above steps need to be repeated until all the required data are collected.

[0088] Afterwards, click on training in the touch display 600, the stored data is processed by the signal fusion algorithm, and then input into the deep learning model as a training set for training to generate training weights;

[0089] Specifically, the data stored in the training area of ​​the edge computing device is transmitted to the embedded data processing algorithm for signal fusion, and the processed data is input into the deep learning network model for training. The weights generated by the training can be used for prediction in the next step, and users can rename, export, delete, and perform other operations on the weights.

[0090] Finally, click Prediction in the software panel, repeat the above data acquisition and training steps, process the newly collected data through the signal fusion algorithm, input it into the prediction algorithm, call the training weights generated by the training step to obtain the prediction results, and display them on the touch display 600.

[0091] Specifically, the data stored in the prediction area of ​​the edge computing device is transmitted to the embedded data processing algorithm for signal fusion, and the processed data is input into the prediction algorithm of the corresponding model. The prediction algorithm will call the specified weights to predict the fault status of the fan. Users can export, delete, and perform other operations on the results.

[0092] In summary, the present invention provides a traction fan fault diagnosis test bench based on multi-source signal fusion, and the traction fan fault diagnosis test bench based on multi-source signal fusion includes: a chassis, a component room and a fan room are isolated and arranged inside the chassis, and the fan room is connected to the component room; a fan assembly, the fan assembly is embedded in the fan room; a data acquisition assembly, the data acquisition assembly includes a vibration sensor, a sound sensor, a temperature sensor and a current sensor arranged in the component room arranged on the fan assembly; an analog assembly, the analog assembly includes a heating mechanism arranged in the component room; a processing assembly, the processing assembly includes a PLC module, the processing assembly is arranged in the component room, and is circuit-connected to the fan assembly, the data acquisition assembly and the analog assembly. The present invention simulates the actual structure and heat dissipation effect of the traction system chassis by setting partitions inside the chassis, and can collect traction fan operation data that is closer to the actual working environment. At the same time, a variety of different types of sensors are set to collect multi-source data. The collected data is formed into multi-source fusion through processing equipment, and cooperates with software such as deep learning networks to learn various signal fault characteristics such as sound, current, and vibration, thereby forming an autonomous diagnosis effect. Compared with the existing technology, the comprehensiveness of signal collection in the process of traction fan fault research can be improved, and the automation of signal transmission, processing and analysis can be improved, thereby assisting researchers to improve the efficiency and accuracy of traction fan fault diagnosis research.

[0093] It should be understood that the application of the present invention is not limited to the above examples. For ordinary technicians in this field, improvements or changes can be made based on the above description. All these improvements and changes should fall within the scope of protection of the claims attached to the present invention.

Claims

1. A traction fan fault diagnosis test bench based on multi-source signal fusion, characterized in that: The traction fan fault diagnosis test bench based on multi-source signal fusion includes: A chassis, wherein a component chamber and a fan chamber are isolatedly arranged inside the chassis, and the fan chamber is connected to the component chamber; A fan assembly, the fan assembly is embedded in the fan chamber; A data acquisition component, the data acquisition component comprising a vibration sensor, a sound sensor arranged on the fan component, a temperature sensor and a current sensor arranged in the component room; A simulation component, the simulation component comprising a heating mechanism disposed in the component chamber; A processing component, wherein the processing component comprises a PLC module, and the processing component is arranged in the component room and is connected to the fan component, the data acquisition component and the simulation component through a circuit.

2. The traction fan fault diagnosis test bench based on multi-source signal fusion according to claim 1 is characterized in that: The component chamber includes a first component chamber and a second component chamber which are arranged on both sides of the fan chamber and communicated with the fan chamber; A plurality of first ventilation holes are arranged on the side wall of the chassis corresponding to the first component chamber; A plurality of second ventilation holes are arranged on the side wall of the chassis corresponding to the second element chamber.

3. The traction fan fault diagnosis test bench based on multi-source signal fusion according to claim 2 is characterized in that: The fan chamber is arranged through two opposite side walls of the chassis; The chassis is detachably provided with a fan chamber baffle, and the fan chamber baffle covers both ends of the fan chamber.

4. The traction fan fault diagnosis test bench based on multi-source signal fusion according to claim 2 is characterized in that: A first air guide port is provided at the bottom of the fan chamber, and a space corresponding to the first air guide port is connected to the first element chamber; The first element chamber is provided with an air duct top plate connected to the side wall of the chassis; A plurality of air duct ribs connected to the air duct top plate and the chassis bottom plate are arranged below the air duct top plate, and hollowed-out guide holes are arranged on the plurality of air duct ribs.

5. The traction fan fault diagnosis test bench based on multi-source signal fusion according to claim 4 is characterized in that: A filter is provided in the first element chamber at a position corresponding to the air duct rib; A plurality of second air guide ports are arranged on the side wall of the second element chamber corresponding to the fan chamber, and a dustproof net is embedded in the second air guide port.

6. The traction fan fault diagnosis test bench based on multi-source signal fusion according to claim 4 is characterized in that: The fan assembly comprises: An outer rotor centrifugal fan, the vibration sensor and the sound sensor are fixedly arranged on the outer rotor centrifugal fan; A frame, wherein the frame is connected to the bottom plate of the fan chamber by bolts, the outer rotor centrifugal fan is fixedly connected to the frame, and the outer rotor centrifugal fan corresponds to the position of the first air guide port.

7. The traction fan fault diagnosis test bench based on multi-source signal fusion according to claim 2 is characterized in that: The heating mechanisms are provided with two, and the two heating mechanisms are fixedly arranged in the first element chamber and the second element chamber respectively; Two temperature sensors are provided, and the two temperature sensors are fixedly disposed in the first element chamber and the second element chamber respectively.

8. The traction fan fault diagnosis test bench based on multi-source signal fusion according to claim 2 is characterized in that: The processing components include: The PLC module is arranged in the first element chamber; An edge calculator, the edge calculator is arranged in the second element chamber, and the edge calculator is connected to the PLC module through a circuit; The data acquisition component also includes a data collector, in which the current sensor is integrated, the vibration sensor, the sound sensor and the temperature sensor are connected to the data collector via a circuit, and the data collector is connected to the edge calculator via a circuit.

9. The traction fan fault diagnosis test bench based on multi-source signal fusion according to claim 8 is characterized in that: A touch display is also embedded on the side wall of the chassis corresponding to the second element chamber. The touch display is connected to the edge calculator through a circuit, and the touch display is used to feed back data information.

10. The traction fan fault diagnosis test bench based on multi-source signal fusion according to claim 2 is characterized in that: The two opposite ends of the chassis corresponding to the first component chamber and the second component chamber are open, and the chassis includes two chassis side panels, which are detachably connected to the open parts at the opposite ends of the chassis.

Images