A mechanical health early warning method and system based on the Internet of Things

Abstract

The present application relates to a mechanical health early warning method and system based on the Internet of Things. A communication network is established through the network access information of each vibration sensor, a communication connection is established based on the network access information of all vibration sensors, and a multi-dimensional detection histogram data array is constructed based on the feedback information of all vibration sensors. Simply put, based on the feedback information of all vibration sensors, three-dimensional feedback information is parsed to determine the working position and three-dimensional virtual space position of all vibration sensors. Using the three-dimensional virtual space position, a multi-dimensional detection histogram data array is constructed. The vibration signal of each vibration sensor is received, and based on the real-time vibration signal, data is filled in the multi-dimensional detection histogram data array to form a histogram data array with a timestamp.

Description

Technical Field

[0001] The present application relates to the technical field of mechanical vibration detection, and in particular to a mechanical health early warning method and system based on the Internet of Things. Background Art

[0002] A key indicator of mechanical health is mechanical fatigue damage. Mechanical fatigue refers to the process by which materials and components, under cyclic stress or cyclic strain, gradually develop localized permanent cumulative damage in one or more locations, leading to cracks or sudden complete fracture after a certain number of cycles. When materials and structures are subjected to repeatedly varying loads, even if the stress value never exceeds the material's strength limit, or even falls below its elastic limit, failure may occur. This phenomenon of material and structural damage under repeated alternating loads is called mechanical fatigue failure.

[0003] In practical engineering applications, mechanical fatigue of mechanical structures is difficult to detect, and mechanical fatigue of mechanical structures can affect the working continuity of the entire mechanical system. Therefore, it is necessary to propose a mechanical health early warning method and system based on the Internet of Things to address the problem of mechanical fatigue being difficult to detect. Summary of the Invention

[0004] Based on this, it is necessary to propose a mechanical health early warning method and system based on the Internet of Things to address the difficult-to-detect defects of mechanical fatigue in mechanical structures.

[0005] This application provides a mechanical health early warning method based on the Internet of Things, including:

[0006] Receive network access information from each vibration sensor;

[0007] Based on the network access information of all vibration sensors, a multi-dimensional detection histogram data array is constructed;

[0008] Receive vibration signals from each vibration sensor;

[0009] Based on the real-time vibration signal, a histogram data array with time stamp is formed;

[0010] Using the system time, multiple histograms with timestamps are formed;

[0011] Feedback information on the replacement of the detection target, and return the network access information received from each vibration sensor until N groups of histograms are received to form samples of the early warning method; the group of histograms includes multiple histograms with timestamps for the same detection target in system time; N is a positive integer;

[0012] Using N groups of histogram data arrays, the machine health warning model to be trained is trained;

[0013] Obtaining a trained machinery health warning model;

[0014] Incorporating the received real-time vibration sensor data into the trained mechanical health early warning model to obtain the mechanical health conclusion of the mechanical system monitored by the vibration sensor;

[0015] Based on the mechanical health conclusions, provide feedback on the maintenance plan for the mechanical system.

[0016] Preferably, the receiving network access information of each vibration sensor includes:

[0017] Select the network access information of a vibration sensor;

[0018] Determine the device serial number of the vibration sensor based on the network access information of the vibration sensor;

[0019] Use the device serial number of the vibration sensor to determine the network communication method of the vibration sensor;

[0020] Return to the network access information of selecting a vibration sensor until all vibration sensors are selected.

[0021] Preferably, the multi-dimensional detection histogram data matrix is ​​constructed based on the network access information of all vibration sensors, including:

[0022] Establishing a virtual three-dimensional position coordinate system for the multi-dimensional detection histogram data array;

[0023] Based on the virtual three-dimensional position coordinate system, the position point of the vibration sensor is set;

[0024] Select a vibration sensor;

[0025] receiving location information of the vibration sensor based on the selected network access communication mode of the vibration sensor;

[0026] Using the position information of the vibration sensor, the position information of the selected vibration sensor is incorporated into the position point of the vibration sensor;

[0027] Return to the step of selecting a vibration sensor until all vibration sensors are selected.

[0028] Preferably, the step of constructing a multi-dimensional detection histogram data matrix based on the network access information of all vibration sensors further includes:

[0029] Saving the three-dimensional data coordinates of the position point of the vibration sensor in the virtual three-dimensional position coordinate system;

[0030] Call a three-dimensional data coordinate;

[0031] Establish vector parameters based on the called three-dimensional data coordinates;

[0032] The calling of a three-dimensional data coordinate is returned until all three-dimensional data coordinates of the position points of the vibration sensor in the virtual three-dimensional position coordinate system are called.

[0033] Preferably, after receiving the vibration signal from each vibration sensor, the method further comprises:

[0034] Select a vibration signal from a vibration sensor;

[0035] Analyze the vibration signal of the selected vibration sensor;

[0036] Obtain the vibration intensity and spatial orientation of the vibration sensor;

[0037] The vibration intensity and spatial orientation of the vibration sensor are used as vector parameters;

[0038] The vibration signal of selecting a vibration sensor is returned until all vibration sensors are selected.

[0039] Preferably, before forming a histogram data array with a time stamp based on the real-time vibration signal, the method includes:

[0040] Call the initial vibration waveform of the initial earthquake source;

[0041] Determine the first-order vibration frequency of the detection target;

[0042] Connect the ends of the vector parameters of the virtual three-dimensional position coordinate system in sequence;

[0043] Obtaining a three-dimensional graph of actual resonance;

[0044] Determine whether the initial vibration waveform is consistent with the actual resonance three-dimensional curve graph;

[0045] If the initial vibration waveform and the actual resonance three-dimensional curve Figure 1 If the mechanical health of the mechanical system meets the threshold,

[0046] If the initial vibration waveform is inconsistent with the actual resonance three-dimensional curve graph, it is determined whether the actual resonance three-dimensional curve graph is consistent with the first-order vibration frequency of the detection target.

[0047] Preferably, the determining whether the actual resonance three-dimensional curve graph is consistent with the first-order vibration frequency of the detection target includes:

[0048] If the actual resonance three-dimensional curve is consistent with the first-order vibration frequency of the detection target, it is determined that the mechanical system is damaged;

[0049] If the actual resonance three-dimensional curve graph is inconsistent with the first-order vibration frequency of the detection target, the instruction feedback of executing the analysis program is fed back; the analysis program is the actual resonance three-dimensional curve graph analysis program.

[0050] Preferably, the parsing procedure includes:

[0051] Call the first-order vibration frequency of the detection target to generate the first-order vibration diagram;

[0052] Using the time stamp, the first-order vibration diagram and the actual resonance three-dimensional curve diagram corresponding to the first-order vibration diagram are matched;

[0053] Determine the same frequency point and asynchronous point;

[0054] The timestamps, the same-frequency points and the asynchronous points are included in the dataset of the histogram with timestamps.

[0055] Preferably, the forming of a histogram data array with a time stamp based on the real-time vibration signal includes:

[0056] Call the dataset that includes timestamps, same-frequency points and asynchronous points into a histogram with timestamps;

[0057] Replace the same-frequency points and asynchronous points of the multi-dimensional detection histogram data array based on the timestamp of the data set;

[0058] Form a histogram of data with timestamps.

[0059] This application also provides a machinery health early warning system based on the Internet of Things, including:

[0060] A host computer, used to execute the mechanical health early warning method based on the Internet of Things;

[0061] The vibration sensors are all communicatively connected with the host computer.

[0062] The present application relates to a mechanical health early warning method and system based on the Internet of Things, which uses the network access information of each vibration sensor to establish a communication network. Based on the network access information of all vibration sensors, a communication connection is established, and based on the feedback information of all vibration sensors, a multi-dimensional detection histogram data array is constructed. Simply put, based on the feedback information of all vibration sensors, three-dimensional feedback information is parsed to determine the working position and three-dimensional virtual space position of all vibration sensors. Using the three-dimensional virtual space position, a multi-dimensional detection histogram data array is constructed. The vibration signal of each vibration sensor is received, and based on the real-time vibration signal, the data of the multi-dimensional detection histogram data array is filled to form a histogram data array with a timestamp. Feedback information on the replacement of the detection target, return to receive the network access information of each vibration sensor, until N groups of histogram data arrays are received, and multiple histogram data arrays with timestamps are formed using the system time.

[0063] A set of histogram data arrays includes multiple timestamped histogram data arrays for the same detection target in system time. The N sets of histogram data arrays are used to train the mechanical health warning model to be trained. The trained mechanical health warning model is then obtained. The received real-time vibration sensor data is incorporated into the trained mechanical health warning model to obtain a conclusion about the mechanical health of the mechanical system monitored by the vibration sensor. Based on the conclusion, a maintenance plan for the mechanical system is provided. This achieves effective detection of mechanical fatigue in mechanical structures. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 A schematic diagram of a method flow of a machine health early warning method based on the Internet of Things provided in one embodiment of the present application.

[0065] Figure 2 This is a structural connection diagram of a machinery health warning system based on the Internet of Things provided in one embodiment of the present application.

[0066] Reference numerals: 100 - host computer; 200 - vibration sensor. DETAILED DESCRIPTION

[0067] In order to make the purpose, technical solutions and advantages of this application more clearly understood, the present application is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0068] This application provides a mechanical health early warning method based on the Internet of Things.

[0069] like Figure 1 As shown, in one embodiment of the present application, a machine health early warning method based on the Internet of Things includes:

[0070] S100, receiving network access information of each vibration sensor;

[0071] S200, constructing a multi-dimensional detection histogram data array based on the network access information of all vibration sensors;

[0072] S300, receiving a vibration signal from each vibration sensor;

[0073] S400, based on the real-time vibration signal, forms a histogram data array with time stamps;

[0074] S500, using the system time, forming multiple histogram data arrays with time stamps;

[0075] S600, feeding back information about the replacement of the detection target, and returning to the process of receiving network access information for each vibration sensor until N sets of histograms are received to form a sample of the early warning method; the set of histograms includes multiple histograms with timestamps for the same detection target in system time; N is a positive integer;

[0076] S700, using N groups of histogram data arrays to train the machine health warning model to be trained;

[0077] Specifically, the method of using N groups of histogram data arrays to train the mechanical health warning model to be trained includes the following steps:

[0078] Based on the multi-dimensional axes of the multi-dimensional detection histogram data array, that is, the identification row factor group of each dimensional axis; the vibration signal of the vibration sensor can be used to clarify the factors of the vibration intensity dimensional axis to be analyzed. For example, in product quality analysis, these factors may include factors such as the function, performance, and appearance of the product, which will serve as the vibration intensity dimensional axis. It can be understood that the column factor group is identified: the factors of the second dimensional axis related to the row factor group are determined. For example, for the above-mentioned product quality analysis example, the column factors can be factors such as production processes, raw materials, and personnel operations that affect product quality. These second dimensional axes can serve as influencing factors of the maintenance plan of the mechanical system when issuing mechanical health warnings.

[0079] In fact, in the process of constructing the mechanical health early warning model, the matrix diagram type can be selected from L-type matrix diagram, T-type matrix diagram, Y-type matrix diagram, and C-type matrix diagram.

[0080] The T-shaped matrix is ​​the most basic matrix diagram, suitable for showing the relationship or correlation between two sets of factors. For example, to analyze the relationship between employee work efficiency and work quality, you can use an L-shaped matrix diagram, with the employee name or number as the row factor and work efficiency and work quality as the column factors.

[0081] Because timestamps are real-time, the interval between inspections during maintenance often determines the suitability of the maintenance plan for that inspection cycle. Simple real-time monitoring often results in strong short-term correlations and excessively high long-term maintenance costs.

[0082] For this reason, the L-type matrix is ​​suitable for depicting the relationship or correlation between two sets of paired factors, making it the most basic matrix diagram. For example, to analyze the relationship between employee work efficiency and work quality, an L-type matrix can be used, with the employee's name or ID as the row factor and work efficiency and work quality as the column factors. Compared to the L-type matrix, the T-type matrix is ​​a combination of an L-type matrix for factors A and B and an L-type matrix for factors A and C. It can be used to analyze the "defective phenomenon - cause - process" relationship in quality issues, or to analyze the relationship between material composition, properties, and application.

[0083] Multi-dimensional detection histograms composed of such rectangular graphs are often feasible solutions for simple maintenance solutions that are efficient, time-efficient, and cost-effective.

[0084] S800, obtains the trained machinery health warning model;

[0085] S900 incorporates the received real-time vibration sensor data into the trained mechanical health early warning model to obtain a conclusion on the mechanical health of the mechanical system monitored by the vibration sensor;

[0086] S910, based on the mechanical health conclusion, provides feedback on the maintenance plan of the mechanical system.

[0087] This embodiment relates to a mechanical health early warning method and system based on the Internet of Things. Communication networking is achieved through the network access information of each vibration sensor. Communication connections are established based on the network access information of all vibration sensors, and a multi-dimensional detection histogram data array is constructed based on the feedback information of all vibration sensors. Simply put, three-dimensional feedback information is parsed based on the feedback information of all vibration sensors to determine the working positions and three-dimensional virtual space positions of all vibration sensors. A multi-dimensional detection histogram data array is constructed using the three-dimensional virtual space positions. Vibration signals are received from each vibration sensor, and data is filled in the multi-dimensional detection histogram data array based on the real-time vibration signals to form a histogram data array with a timestamp. Information about the replacement of the detection target is fed back, and the network access information of each vibration sensor is returned until N groups of histogram data arrays are received. Multiple histogram data arrays with timestamps are formed using the system time.

[0088] A set of histogram data arrays includes multiple timestamped histogram data arrays for the same detection target in system time. The N sets of histogram data arrays are used to train the mechanical health warning model to be trained. The trained mechanical health warning model is then obtained. The received real-time vibration sensor data is incorporated into the trained mechanical health warning model to obtain a conclusion about the mechanical health of the mechanical system monitored by the vibration sensor. Based on the conclusion, a maintenance plan for the mechanical system is provided. This achieves effective detection of mechanical fatigue in mechanical structures.

[0089] In one embodiment of the present application, S100 includes:

[0090] S110, selecting network access information of a vibration sensor;

[0091] Specifically, computer networking based on the host computer connects multiple computers or other network devices to achieve a complex process of data exchange and resource sharing;

[0092] Simply put, clarify the scale of the network, whether it is a small office, home network, or a large enterprise network. Determine the scope of the network coverage, such as a local area network (LAN) is suitable for a small geographical area, and a wide area network (WAN) is used for inter-regional connections;

[0093] Choose a network topology, such as bus topology, star topology, ring topology, tree topology, and mesh topology. In bus topology, all devices are connected through a main line. It has a simple structure, but low reliability. A node failure may affect the entire network. It is often used in small networks. Star topology uses a switch or hub as the central node, and other devices are connected to it through network cables. It is easy to manage and maintain, has strong scalability, and is one of the most common networking methods. Ring topology connects devices in sequence to form a closed loop. Signals are transmitted unidirectionally in the ring. Data transmission is stable, but reliability is poor. A node failure may cause the entire network to be paralyzed. Tree topology combines the characteristics of bus and star topologies. It has a hierarchical structure and is suitable for large networks. It has good scalability, but relatively complex maintenance. The mesh topology directly connects each device to multiple other devices. It has extremely high reliability and redundancy, but the cost is high. It is often used in situations where network reliability is extremely high.

[0094] S120, determining a device serial number of the vibration sensor based on the network access information of the vibration sensor;

[0095] Specifically, the vibration sensor, as a complete device, can implement the parsing of the vibration sensor's network access information based on the modem to achieve the complex process of data exchange and resource sharing;

[0096] The parsed device serial number of the vibration sensor can be used as the network access identifier of the vibration sensor;

[0097] S130: Determine a network access communication mode of the vibration sensor using the device serial number of the vibration sensor.

[0098] Specifically, different types of vibration sensors have different working processes. For IoT mechanical health warning vibration sensors, the installation of complete device vibration sensors often affects the vibration damping of the entire mechanical system.

[0099] When the installation of the vibration sensor of the complete device does not affect the vibration damping of the entire mechanical system, the vibration signal of the vibration sensor can be used to clarify the factors of the vibration intensity dimension axis to be analyzed. For example, in product quality analysis, factors such as product function, performance, and appearance may be included. These factors will serve as the vibration intensity dimension axis. It can be understood that the column factor group is identified: the factors of the second dimension axis related to the row factor group are determined. For example, for the above product quality analysis example, the column factors can be factors such as production process, raw materials, and personnel operations that affect product quality. These second-dimensional axes can serve as influencing factors of the maintenance plan of the mechanical system during mechanical health warning.

[0100] S140, returning the network access information of selecting a vibration sensor until all vibration sensors are selected.

[0101] Specifically, computer networking is a systematic project. This networking process meets the needs of long-term monitoring. When considering multiple factors, the computer networking should be based on the specific use requirements of the mechanical equipment, such as usage time, installation level, work intensity, etc. Based on the factor of work intensity alone, the selection of appropriate networking methods and technical means ensures the efficiency of network communication, the reliability and security of detection and early warning;

[0102] Simply put, when a vibration sensor fails, long-term monitoring data can have a higher early warning effect.

[0103] In one embodiment of the present application, S200 includes:

[0104] S210, establishing a virtual three-dimensional position coordinate system of the multi-dimensional detection histogram data array;

[0105] S220, setting a position point of the vibration sensor based on the virtual three-dimensional position coordinate system;

[0106] S230, selecting a vibration sensor;

[0107] S240, receiving location information of the vibration sensor based on the selected network access communication mode of the vibration sensor;

[0108] S250, using the position information of the vibration sensor, incorporating the position information of the selected vibration sensor into the position point of the vibration sensor;

[0109] S260, returning to the step of selecting a vibration sensor until all vibration sensors are selected;

[0110] Specifically, determine the type of vibration sensor, such as vibration intensity sensor, vibration vector sensor, and vibration variance sensor. Different virtual space application scenarios have different requirements for the coordinate system. For example, if it is used for virtual game development, a coordinate system that can adapt to various terrains and game mechanics may be required. When mechanical structures are used to simulate the virtual vibration environment of game development, vibration vector sensors based on vibration intensity sensors and vibration variance sensors can provide mechanical health warnings based on the Internet of Things.

[0111] Real-time monitoring data from mechanical structures can simply depict the virtual vibration environment used in simulated game development. While the host computer often has significant data redundancy when utilizing the Internet of Things (IoT), processing this redundant data can be simplified by using multi-dimensional histogram data arrays. This simplifies environmental deployment issues such as the size, proportions, and spatial relationships of the mechanical structure. Environmental deployment is crucial for mechanical systems. Rust, oil film, and other conditions in mechanical systems can affect vibration transmission. Data simulation, modeling, and early warning can easily improve the accuracy of real-time monitoring and early warning.

[0112] In one embodiment of the present application, S200 further includes:

[0113] S270, saving the three-dimensional data coordinates of the position point of the vibration sensor in the virtual three-dimensional position coordinate system;

[0114] S280, calling a three-dimensional data coordinate;

[0115] S290, establishing vector parameters based on the called three-dimensional data coordinates;

[0116] S281, returning to the calling of a three-dimensional data coordinate, until all three-dimensional data coordinates of the position points of the vibration sensor in the virtual three-dimensional position coordinate system are called.

[0117] Specifically, by introducing additional coordinates, points in n-dimensional space are represented as n+1-dimensional vectors. Homogeneous coordinates are often used in computer graphics to represent two-dimensional and three-dimensional points, facilitating various transformations such as translation, rotation, and scaling. For virtual space scenes requiring complex graphical transformations, such as animation and object manipulation in virtual reality, homogeneous coordinate systems can provide a more concise and unified mathematical description.

[0118] In one embodiment of the present application, after S100, the following steps are included:

[0119] S150, selecting a vibration signal from a vibration sensor;

[0120] S160, analyzing the vibration signal of the selected vibration sensor;

[0121] S170, obtaining the vibration intensity and spatial orientation of the vibration sensor;

[0122] S180, taking the vibration intensity and spatial direction of the vibration sensor as vector parameters;

[0123] S190, returning to the process of selecting a vibration signal of a vibration sensor until all vibration sensors are selected.

[0124] The spatial orientation of vibration describes the behavior and propagation characteristics of vibration in a specific direction. It is crucial for understanding vibration phenomena and analyzing vibration systems. Spatial orientation refers to the directionality of vibration in three-dimensional space, specifically the distribution and propagation of vibration energy in different directions. It can be described by the components of physical quantities such as displacement, velocity, or acceleration in each spatial direction. Typically, in a three-dimensional Cartesian coordinate system, vibration components along the x, y, and z axes are analyzed. The x-axis generally represents horizontal or front-to-back direction, the y-axis represents vertical or left-to-right direction, and the z-axis represents up-down direction. By measuring changes in displacement, velocity, or acceleration of an object along these coordinate axes, the spatial orientation of vibration can be determined. Different vibration sources produce vibrations with different spatial orientations. For example, rotating machinery, due to its structural characteristics and motion, may produce vibrations that are more pronounced in certain directions.

[0125] In one embodiment of the present application, before S400, the following steps are included:

[0126] S410, calling the initial vibration waveform of the initial earthquake source;

[0127] S420, determining the first-order vibration frequency of the detection target;

[0128] S430, sequentially connecting the ends of the vector parameters of the virtual three-dimensional position coordinate system;

[0129] S440, obtaining a three-dimensional graph of actual resonance.

[0130] In one embodiment of the present application, S400 includes:

[0131] S450, determining whether the initial vibration waveform is consistent with the actual resonance three-dimensional curve graph;

[0132] S460, if the initial vibration waveform and the actual resonance three-dimensional curve Figure 1 If the mechanical health of the mechanical system meets the threshold,

[0133] S470, if the initial vibration waveform is inconsistent with the actual resonance three-dimensional curve graph, determining whether the actual resonance three-dimensional curve graph is consistent with the first-order vibration frequency of the detection target;

[0134] S480, if the actual resonance three-dimensional curve graph is consistent with the first-order vibration frequency of the detection target, then it is determined that the mechanical system is damaged;

[0135] S490, if the actual resonance three-dimensional curve graph is inconsistent with the first-order vibration frequency of the detection target, feeding back an instruction feedback for executing an analysis program; the analysis program is an actual resonance three-dimensional curve graph analysis program;

[0136] S491, calling a first-order vibration graph generated by the first-order vibration frequency of the detection target;

[0137] S492, using the timestamp, matching the first-order vibration diagram and the actual resonance three-dimensional curve diagram corresponding to the first-order vibration diagram;

[0138] S493, determine the co-frequency point and asynchronous point;

[0139] S494, incorporate the timestamp, the co-frequency points, and the asynchronous points into a data set of a histogram with the timestamp.

[0140] In one embodiment of the present application, S400 further includes:

[0141] K100, calls the dataset that includes timestamps, same-frequency points, and asynchronous points into a histogram with timestamps;

[0142] K200 replaces the same-frequency points and asynchronous points of the multi-dimensional detection histogram data array based on the timestamp of the data set;

[0143] K300, forming a histogram data array with time stamps.

[0144] Specifically, timestamps, co-frequency points, and asynchronous points can form different time periods, thereby simply dividing the monitoring interval and improving the accuracy of early warnings.

[0145] This application provides a machinery health early warning system based on the Internet of Things.

[0146] like Figure 2 As shown, in one embodiment of the present application, a mechanical health warning system based on the Internet of Things includes a host computer 100 and a vibration sensor 200. The host computer 100 is used to execute a mechanical health warning method based on the Internet of Things, and the vibration sensor 200 is communicatively connected to the host computer 100.

[0147] This embodiment relates to a mechanical health early warning system based on the Internet of Things. The host computer 100 establishes a communication network using the network access information of each vibration sensor 200. A communication connection is established based on the network access information of all vibration sensors 200, and a multidimensional detection histogram data matrix is ​​constructed based on the feedback information of all vibration sensors 200. Simply put, based on the feedback information of all vibration sensors 200, three-dimensional feedback information is parsed to determine the working position and three-dimensional virtual space position of all vibration sensors 200. The three-dimensional virtual space position is used to construct a multidimensional detection histogram data matrix. A vibration signal is received from each vibration sensor 200, and data is filled in the multidimensional detection histogram data matrix based on the real-time vibration signal to form a histogram data matrix with a timestamp. Information about the change of detection target is fed back, and the network access information of each vibration sensor 200 is returned until N groups of histogram data matrixes are received. Multiple histogram data matrixes with timestamps are formed using the system time.

[0148] A set of histogram data arrays includes multiple timestamped histogram data arrays for the same detection target in system time. The N sets of histogram data arrays are used to train the mechanical health warning model to be trained. The trained mechanical health warning model is then obtained. The received real-time vibration sensor data is incorporated into the trained mechanical health warning model to obtain a conclusion about the mechanical health of the mechanical system monitored by the vibration sensor. Based on the conclusion, a maintenance plan for the mechanical system is provided. This achieves effective detection of mechanical fatigue in mechanical structures.

[0149] The various technical features of the above-described embodiments can be combined arbitrarily, and the execution order of the method steps is not restricted. In order to make the description concise, not all possible combinations of the various technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0150] The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present application. It should be noted that a person of ordinary skill in the art may make various modifications and improvements without departing from the spirit of the present application, and these modifications and improvements fall within the scope of protection of the present application. Therefore, the scope of protection of the present application shall be determined by the appended claims.

Claims

1. A mechanical health early warning method based on the Internet of Things, characterized in that: include: Receive network access information from each vibration sensor; Based on the network access information of all vibration sensors, a multi-dimensional detection histogram data array is constructed; Receive vibration signals from each vibration sensor; Based on the real-time vibration signal, a histogram data array with time stamp is formed; Using the system time, multiple histograms with timestamps are formed; Feedback information on the replacement of the detection target, and return the network access information received from each vibration sensor until N groups of histograms are received to form a sample of the early warning method; the group of histograms includes multiple histograms with timestamps for the same detection target in the system time; N is a positive integer; Using N groups of histogram data arrays, the machine health warning model to be trained is trained; Obtaining a trained machinery health warning model; Incorporating the received real-time vibration sensor data into the trained mechanical health early warning model to obtain the mechanical health conclusion of the mechanical system monitored by the vibration sensor; Provide maintenance plan for the mechanical system based on the conclusion of mechanical health; The multi-dimensional detection histogram data matrix is ​​constructed based on the network access information of all vibration sensors, including: Establishing a virtual three-dimensional position coordinate system for the multi-dimensional detection histogram data array; Based on the virtual three-dimensional position coordinate system, the position point of the vibration sensor is set; Select a vibration sensor; receiving location information of the vibration sensor based on the selected network access communication mode of the vibration sensor; Using the position information of the vibration sensor, the position information of the selected vibration sensor is incorporated into the position point of the vibration sensor; Return to the step of selecting a vibration sensor until all vibration sensors are selected; Saving the three-dimensional data coordinates of the position point of the vibration sensor in the virtual three-dimensional position coordinate system; Call a three-dimensional data coordinate; Establish vector parameters based on the called three-dimensional data coordinates; Returning the calling of a three-dimensional data coordinate until all three-dimensional data coordinates of the position points of the vibration sensor in the virtual three-dimensional position coordinate system are called; Before forming a histogram data array with a time stamp based on the real-time vibration signal, the method includes: Call the initial vibration waveform of the initial earthquake source; Determine the first-order vibration frequency of the detection target; Connect the ends of the vector parameters of the virtual three-dimensional position coordinate system in sequence; Obtaining a three-dimensional graph of actual resonance; Determine whether the initial vibration waveform is consistent with the actual resonance three-dimensional curve graph; If the initial vibration waveform is consistent with the actual resonance three-dimensional curve graph, it is determined that the mechanical health of the mechanical system meets the threshold; If the initial vibration waveform is inconsistent with the actual resonance three-dimensional curve graph, then determine whether the actual resonance three-dimensional curve graph is consistent with the first-order vibration frequency of the detection target; The determining whether the actual resonance three-dimensional curve graph is consistent with the first-order vibration frequency of the detection target includes: If the actual resonance three-dimensional curve is consistent with the first-order vibration frequency of the detection target, it is determined that the mechanical system is damaged; If the actual resonance three-dimensional curve graph is inconsistent with the first-order vibration frequency of the detection target, the instruction feedback of executing the analysis program is fed back; the analysis program is the actual resonance three-dimensional curve graph analysis program.

2. The machine health early warning method based on the Internet of Things according to claim 1 is characterized in that: The receiving of network access information of each vibration sensor includes: Select the network access information of a vibration sensor; Determine the device serial number of the vibration sensor based on the network access information of the vibration sensor; Use the device serial number of the vibration sensor to determine the network communication method of the vibration sensor; Return to the network access information of selecting a vibration sensor until all vibration sensors are selected.

3. The machine health early warning method based on the Internet of Things according to claim 1 is characterized in that: After receiving the vibration signal from each vibration sensor, the method includes: Select a vibration signal from a vibration sensor; Analyze the vibration signal of the selected vibration sensor; Obtain the vibration intensity and spatial orientation of the vibration sensor; The vibration intensity and spatial orientation of the vibration sensor are used as vector parameters; The vibration signal of selecting a vibration sensor is returned until all vibration sensors are selected.

4. The machine health early warning method based on the Internet of Things according to claim 1 is characterized in that: The parsing procedure includes: Call the first-order vibration frequency of the detection target to generate the first-order vibration diagram; Using the time stamp, the first-order vibration diagram and the actual resonance three-dimensional curve diagram corresponding to the first-order vibration diagram are matched; Determine the same frequency point and asynchronous point; The timestamps, the same-frequency points and the asynchronous points are included in the dataset of the histogram with timestamps.

5. The machine health early warning method based on the Internet of Things according to claim 4 is characterized in that: The histogram data array with time stamp is formed based on the real-time vibration signal, including: Call the dataset that includes timestamps, same-frequency points and asynchronous points into a histogram with timestamps; Replace the same-frequency points and asynchronous points of the multi-dimensional detection histogram data array based on the timestamp of the data set; Form a histogram of data with timestamps.

6. A mechanical health early warning system based on the Internet of Things, characterized in that: include: A host computer, configured to execute the machine health early warning method based on the Internet of Things according to any one of claims 1 to 5; The vibration sensors are all connected to the host computer for communication.

Images