A digital triplet system construction method for complex equipment fault diagnosis

Abstract

The present invention discloses a method for constructing a digital triplet system for complex equipment fault diagnosis. The method comprises the following steps: S1 six-step simplification: completing the construction of a semi-physical entity of complex equipment by simplifying the entity of complex equipment in six steps; S2 architecture construction: creating entities, the connection relationship between the semi-physical entity and the virtual entity, and completing the structure and overall connection of the digital triplet; S3 fault simulation: performing fault simulation and creating fault types on the semi-physical entity end of the digital triplet system architecture, and constructing a complex equipment fault data set; S4 practical application: mapping from the semi-physical entity end to the physical entity end, thereby performing status monitoring and fault diagnosis on the physical entity end. The present invention can effectively solve the problem that faults cannot be simulated and verified under the digital twin architecture, performs fault simulation and verification based on the digital triplet system, improves the efficiency of complex equipment fault diagnosis, and greatly promotes the combination of digital twin technology and fault diagnosis technology.

Description

Technical Field

[0001] The present invention belongs to the field of complex equipment digital twin construction, and in particular relates to a digital triplet system construction method for complex equipment fault diagnosis. Background Art

[0002] The continuous and in-depth integration of informatization and industrialization has promoted the development of intelligent manufacturing and the application of digital twin technology in industrial production. Digital twin technology can effectively participate in all stages of the production process, improving production safety. The operation and maintenance stage, a critical part of the industrial production process, relies on sensors for equipment monitoring. Digital twin-driven fault diagnosis can transform post-event maintenance into predictive maintenance, thereby improving the efficiency and safety of the operation and maintenance process. However, due to the complex coupling between parts of complex equipment and the inflexible deployment of physical sensors, data collection and fault analysis are difficult. Therefore, it is impossible to effectively correlate fault characteristics in the digital twin. Studying fault mechanisms solely from the physical space of complex equipment is time-consuming, labor-intensive, and costly. Failures in complex equipment pose a significant threat to public safety. Therefore, based on the current challenges of digital twin fault diagnosis, a digital triplet system construction method for complex equipment fault diagnosis is developed. This method effectively addresses the current problem of fault diagnosis being unable to conduct two-way simulation verification. It also allows for flexible sensor deployment for data collection, improves fault diagnosis efficiency, and promotes the integration of digital twins and fault diagnosis technology. Summary of the Invention

[0003] In order to solve the problems in the background technology, the present invention proposes a method for constructing a digital triplet system for complex equipment fault diagnosis. The digital triplet is optimized on the digital twin infrastructure, and the constructed complex equipment entity makes the digital triplet system more suitable for complex equipment fault diagnosis.

[0004] The present invention can effectively solve the problem that faults cannot be simulated and verified under the digital twin architecture. By constructing the physical end, the virtual physical end and the semi-physical physical end, the digital triplet system is built, and fault simulation and verification are performed based on the digital triplet system, which improves the efficiency of fault diagnosis of complex equipment. It is highly innovative and greatly promotes the combination of digital twin technology and fault diagnosis technology.

[0005] The technical solution adopted by the present invention comprises the following steps:

[0006] S1: Simplify the physical entity of complex equipment in six steps to complete the construction of the semi-physical entity of complex equipment; the six steps of simplification include functional division, structural simplification, similarity contraction, structural optimization, structural definition and system integration;

[0007] The complex equipment includes high-speed elevators, escalators, offshore wind turbines, ships, and high-speed trains;

[0008] S2: Build a digital triplet system for complex equipment to realize the connection between the physical entity, semi-physical entity and virtual entity of complex equipment;

[0009] S3: Fault simulation based on semi-physical entities of complex equipment and collection of fault data;

[0010] S4: Map from the semi-physical entity end of complex equipment to the physical entity end of complex equipment, so as to perform fault diagnosis on the physical entity of complex equipment and visualize it in the virtual end of complex equipment.

[0011] The step S1 is specifically as follows:

[0012] The construction from complex equipment physical entity to complex equipment semi-physical entity mainly includes six steps of simplification process: function division, structure simplification, similarity contraction, structure optimization, structure definition and system integration;

[0013] Step 1.1) Functional division: Divide the complex equipment into multiple systems, i.e., multiple functional modules, according to its structural functions;

[0014] Step 1.2) Simplify the structure: Identify the relevant systems that affect the occurrence of complex equipment failures and make them the main functional systems. Delete all other systems except the main functional systems.

[0015] Step 1.3) Similarity contraction:

[0016] Determine the ratio of physical entities and semi-physical entities, and perform similarity contraction on the physical entities of complex equipment based on the principle of similarity, where similarity includes geometric, kinematic and dynamic similarity;

[0017] Determine the dimensions of components in the main functional system according to the set ratio, create the corresponding CAD model, and obtain the model parts library for unified management of complex equipment parts;

[0018] Step 1.4) Structural optimization: Delete components in the main functional system that are not related to the fault;

[0019] Step 1.5) Structure definition: Match the corresponding components in the CAD model with the corresponding components of the physical entity functional module;

[0020] Step 1.6) System Integration: Process the non-standard parts formed by structural simplification, similarity contraction and structural optimization according to the CAD model, and purchase the standard parts required by the CAD model in the parts library. Reassemble the non-standard parts and standard parts to integrate them into a semi-physical entity of the complex equipment.

[0021] Perform similarity verification on the complex semi-physical entity of the equipment integrated in step 1.6), specifically:

[0022] Based on the Buckingπ theorem, the dimensionless representation of the semi-physical entity and the physical entity is obtained. According to the actual size of the physical entity and the custom size of the semi-physical entity, the dimensionless ratio of the physical entity and the dimensionless ratio of the semi-physical entity are calculated respectively, and the dimensionless ratios of the two are compared:

[0023] A threshold value is set, and the specific size values ​​of the physical entity and the semi-physical entity are substituted into the dimensionless value to calculate the dimensionless value. When the difference between the dimensionless ratios of the two does not exceed the threshold, the semi-physical entity formed in step 1.6) is used as the final semi-physical entity of the complex equipment. When the difference between the dimensionless ratios of the two exceeds the threshold, the process returns to the structural optimization in step 1.4) and resizes the relevant parts in the parts library until the difference between the dimensionless ratios of the two meets the threshold requirement.

[0024] The digital triplet system construction in step S2 is specifically as follows:

[0025] Use visualization software to create a virtual entity with the same size as the semi-physical entity; this application uses Unity software

[0026] Acceleration sensors are placed on both the physical entity and the semi-physical entity. Data collected by the physical entity and the semi-physical entity are transmitted to the virtual entity in real time through the sensors. The virtual entity visualizes the received data and stores the data in a database.

[0027] Programmable logic controllers are installed on both the physical entity and the semi-physical entity. The virtual entity transmits commands to the controller of the physical entity or the semi-physical entity, and adjusts the operating status of the physical entity or the semi-physical entity through the controller.

[0028] The step S3 is specifically as follows:

[0029] For components that need to be investigated during the operation of complex equipment or components that often exhibit abnormalities, matching components are found in the semi-physical entity of the complex equipment according to the structural definition in step S1. Damage simulation is performed on the matching components, such as tension, compression, torsion, and physical deformation in mechanical damage, and different labels are assigned to different damage types.

[0030] Sensors are used to collect data from semi-physical entities after damage simulation, and form a complex equipment fault data set.

[0031] The step S4 is specifically as follows:

[0032] The diagnostic model is trained using the fault data set of the semi-physical entity obtained in step S3, and the trained diagnostic model is actually applied in the physical entity of the complex equipment. The training process and diagnostic results are visualized in the virtual entity of the complex equipment.

[0033] The diagnostic model includes a convolutional neural network model

[0034] Since the semi-physical entity is similar to the physical entity, the diagnostic methods that can be implemented on the semi-physical entity are also applicable to the physical entity.

[0035] Beneficial effects of the present invention:

[0036] The present invention simplifies the physical entity of complex equipment to form a semi-physical entity of complex equipment. Based on the semi-physical entity of complex equipment, sensors can be flexibly arranged to collect data, simulate and verify faults, and complete the mapping between the complex equipment entity. The virtual end of the complex equipment visualizes the operating status of the complex equipment to realize the monitoring of the complex equipment.

[0037] Compared with the existing technology, the present invention can effectively solve the problem that the current digital twin-driven fault diagnosis cannot achieve a two-way closed loop of simulation verification, and promote the combination of digital twin technology and fault diagnosis technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 Schematic diagram of the process of the present invention.

[0039] Figure 2 This is the corresponding diagram of the six-step simplified process system.

[0040] Figure 3 This is the visualization result of the elevator virtual entity.

[0041] Figure 4 This is a semi-physical entity fault simulation category for elevators. DETAILED DESCRIPTION

[0042] The specific embodiments of the present invention are described below to facilitate understanding of the present invention by those skilled in the art. However, it should be clear that the present invention is not limited to the scope of the specific embodiments. For those skilled in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the appended claims, these changes are obvious, and all inventions and creations utilizing the concepts of the present invention are protected.

[0043] This paper uses the KLK2 high-speed elevator as an example of complex equipment to illustrate the complete construction and fault simulation of the digital triplet system.

[0044] like Figure 1 As shown, the present invention includes the following steps:

[0045] S1: Complete the semi-physical construction of the high-speed elevator by simplifying the high-speed elevator entity in six steps.

[0046] The process of transforming a physical elevator into a semi-physical elevator involves six steps: functional division, structural simplification, similarity reduction, structural optimization, structural definition, and system integration. To facilitate the application of digital twin technology in elevator fault diagnosis, high-speed elevators are divided into eight functional modules based on their structural functions: the guidance system, traction system, door system, elevator car system, weight and balance system, power system, power control system, and safety system. To ensure system operation, the structure is simplified according to the divided functional modules, and components that have little impact on the functional modules are removed.

[0047] The factors that affect the horizontal vibration of a high-speed elevator car primarily include the guide system and the car system. These systems are considered the primary functional systems, and all other systems are removed. The following decomposition of the high-speed elevator system demonstrates the relationship between the subsystems and the overall system. Assume the system has m physical quantities, n of which are fundamental physical dimensions. This means that mn physical quantities can be expressed using n fundamental physical dimensions:

[0048] f(x1,x2,...,x n |x n+1 ,x n+2 ,...,x m )=0 (1)

[0049] Since the system is a physical system, it can be described by parameters. The overall characteristic equation is equivalent to the sum of the characteristic equations of each subsystem, which can be expressed as follows:

[0050]

[0051] Express (2) in the form of (1):

[0052]

[0053] Since mn physical quantities can be expressed by n basic physical dimensions, then f(G S ) and f(D S ) can be expressed as:

[0054]

[0055]

[0056] By constructing a new basic dimension, we can obtain:

[0057]

[0058] make:

[0059]

[0060] Then we can get:

[0061]

[0062] That is, the right side must be dimensionless like the left side, so the guide system and the car system that affect the horizontal vibration of the car can be analyzed dimensionally separately.

[0063] Construct a car horizontal vibration dynamics model for elevator operation:

[0064]

[0065] in, are the spring force and damping generated by the rolling guide shoe, k and c are the equivalent stiffness and damping. i is the offset of the guide shoe at the installation position. h1h2 are the longitudinal distances from different guide shoes to the center of mass O of the car.

[0066]

[0067] z i =xh i θ-x i

[0068] Through (8-9) we can get

[0069]

[0070] Based on (9), it can be concluded that the factors affecting the horizontal vibration of the elevator are mainly the following physical quantities.

[0071]

[0072] Therefore, the main influencing factors for similar shrinkage can be expressed as

[0073] ele=f(m,J,k,c,y,θ,h1,h2)(11)

[0074] Apply Bucking Π theorem to perform similarity design and obtain

[0075]

[0076] The main parameters of KLK2 are:

[0077]

[0078]

[0079] Formula (7) proves that the elevator guide shoe system and the car system can be analyzed for similarity separately. Therefore, the geometric ratio size is designed to be 10:1 with the guide rail as the reference length, and the car mass ratio is 1000:1. According to (12), the semi-physical entity design size of the high-speed elevator after similar contraction can be obtained as follows:

[0080]

[0081] The relationship between Π1 and Π2 is Substituting the actual dimensions, the high-speed elevator semi-physical design parameters were found to be essentially identical to the physical design parameters: 567.332 and 565.731, respectively. The difference was 1.601, less than the set threshold of 3, meeting the similarity contraction requirement. A high-speed elevator semi-physical model was then constructed based on the design parameters.

[0082] like Figure 2 As shown in the figure, it is the system corresponding diagram after the elevator physical entity is simplified into the elevator semi-physical entity through a six-step process. 2-1 represents the CAD model of the high-speed elevator physical entity, 2-2 represents the CAD model of the high-speed elevator semi-physical entity, and 2-3 represents the high-speed elevator semi-physical entity. Among them, 1 represents the elevator power system, 2 represents the elevator weight balance system, 3 represents the elevator traction system, 4 represents the elevator guide system, and 5 represents the elevator car system.

[0083] S2: The architecture completes the structure and overall connection of the digital triplets.

[0084] A high-speed elevator digital twin system is constructed for high-speed elevators, and a high-speed elevator semi-physical entity is added to the high-speed elevator digital twin system architecture to form a new digital triplet six-dimensional structure. The six dimensions include the elevator physical entity, the elevator semi-physical entity, the elevator virtual entity, the connection between the three entities, the data between the three entities, and the server, such as Figure 1 As shown. The physical entity of the elevator exists in the real space and relies on sensor technology to collect data. The semi-physical entity of the elevator is simplified through the S1 step, and status monitoring is achieved through sensor technology, and it is controlled using an application. The virtual entity of the elevator monitors the high-speed elevator entity and semi-physical entity through modeling and visualization technology. Different entities are connected through communication interfaces, and different communication protocols are used to support data transmission. In this case, the OPC UA protocol is used to interact with data and control devices. The server inputs data generated from the elevator physical entity, the elevator semi-physical entity, and the elevator virtual entity for unified management and analysis, and sends control instructions to different entities through the protocol to control the operating status. This completes the construction of the high-speed elevator digital triplet system. Server-side visualization is implemented through unity, as shown Figure 3The figure shows the elevator virtual entity created by Unity, which corresponds to the elevator semi-physical entity. The virtual entity can be monitored and controlled and fed back to the elevator semi-physical entity to achieve real-time data transmission. Figure 3 The vibration curve of the system is shown in Figure .

[0085] S3: Perform fault simulation and create fault types on the semi-physical entity side of the digital triplet system architecture to build a complex equipment fault dataset

[0086] In the semi-physical entity of the elevator, damage simulation is performed on the components that often experience abnormalities during the operation of high-speed elevators, and labels are assigned to the damage to build a fault dataset, such as Figure 4 As shown, there are five different label categories: normal guide rail 4-1, guide rail step 4-2, guide rail misalignment 4-3, normal roller 4-4, and roller wear 4-5. Sensors are deployed in the high-speed elevator semi-physical entity to collect vibration data for different fault categories, allowing for subsequent diagnostic analysis.

[0087] S4. Perform status monitoring and fault diagnosis on physical entities

[0088] By constructing a diagnostic model, a fault data set constructed on the semi-physical entity of the elevator is diagnosed and analyzed. Since the similarity between the semi-physical entity and the physical entity of the high-speed elevator is verified in S1, the diagnostic model can be applied to the physical entity of the high-speed elevator. Through data transmission, visualization and real-time data transmission are performed on the virtual entity of the high-speed elevator, realizing the overall construction of a digital triplet system for high-speed elevator fault diagnosis.

[0089] The necessity of building digital triplets of complex equipment can be illustrated by the example of high-speed elevators.

[0090] In summary, the present invention proposes a method for constructing a digital triplet system for complex equipment fault diagnosis.

Claims

1. A method for constructing a digital triplet system for complex equipment fault diagnosis, characterized in that: The following steps are involved: S1: Simplify the physical entity of complex equipment in six steps to complete the construction of the semi-physical entity of complex equipment; the six steps of simplification include functional division, structural simplification, similarity contraction, structural optimization, structural definition and system integration; S2: Build a digital triplet system for complex equipment to realize the connection between the physical entity, semi-physical entity and virtual entity of complex equipment; S3: Fault simulation based on semi-physical entities of complex equipment and collection of fault data; S4: Fault diagnosis of complex equipment physical entities and visualization in the complex equipment virtual terminal; The step S1 is specifically as follows: Step 1.1) Functional division: Divide the complex equipment into multiple systems, i.e., multiple functional modules, according to its structural functions; Step 1.2) Simplify the structure: Identify the relevant systems that affect the occurrence of complex equipment failures and make them the main functional systems. Delete all other systems except the main functional systems. Step 1.3) Similarity contraction: Determine the ratio of physical entities and semi-physical entities, and perform similarity contraction on the physical entities of complex equipment based on the principle of similarity, where similarity includes geometric, kinematic and dynamic similarity; Determine the dimensions of components in the main functional system according to the set ratio, create the corresponding CAD model, and obtain the model parts library for unified management of complex equipment parts; Step 1.4) Structural optimization: Delete components in the main functional system that are not related to the fault; Step 1.5) Structure definition: Match the corresponding components in the CAD model with the corresponding components of the physical entity functional module; Step 1.6) System Integration: Process the non-standard parts formed by structural simplification, similarity contraction and structural optimization according to the CAD model, and purchase the standard parts required by the CAD model in the parts library. Reassemble the non-standard parts and standard parts to integrate them into a semi-physical entity of the complex equipment.

2. The method for constructing a digital triplet system for complex equipment fault diagnosis according to claim 1, characterized in that: The similarity verification of the complex equipment semi-physical entity integrated in step 1.6) is performed as follows: Based on the Buckingπ theorem, the dimensionless representation of the semi-physical entity and the physical entity is obtained. According to the actual size of the physical entity and the custom size of the semi-physical entity, the dimensionless ratio of the physical entity and the dimensionless ratio of the semi-physical entity are calculated respectively, and the dimensionless ratios of the two are compared: A threshold value is set. When the difference between the dimensionless ratios of the two does not exceed the threshold value, the semi-physical entity formed in step 1.6) is used as the final semi-physical entity of the complex equipment. When the difference between the dimensionless ratios of the two exceeds the threshold, return to the structural optimization of step 1.4) and modify the dimensions of the relevant parts in the parts library until the difference between the dimensionless ratios of the two meets the threshold requirement.

3. The method for constructing a digital triplet system for complex equipment fault diagnosis according to claim 1, characterized in that: The digital triplet system construction in step S2 is specifically as follows: Creating a virtual entity with the same dimensions as the semi-physical entity through visualization software; Acceleration sensors are placed on both the physical entity and the semi-physical entity. Data collected by the physical entity and the semi-physical entity are transmitted to the virtual entity in real time through the sensors. The virtual entity visualizes the received data and stores the data in a database. Programmable logic controllers are installed on both the physical entity and the semi-physical entity. The virtual entity transmits commands to the controller of the physical entity or the semi-physical entity, and adjusts the operating status of the physical entity or the semi-physical entity through the controller.

4. The method for constructing a digital triplet system for complex equipment fault diagnosis according to claim 1, characterized in that: The step S3 is specifically as follows: For components that need to be explored during the operation of complex equipment, matching components are searched in the semi-physical entity of the complex equipment according to the structure definition in step S1, damage simulation is performed on the matching components, and different labels are assigned to different damage types; Sensors are used to collect data from semi-physical entities after damage simulation, and form a complex equipment fault data set.

5. The method for constructing a digital triplet system for complex equipment fault diagnosis according to claim 1, characterized in that: The step S4 is specifically as follows: The diagnostic model is trained using the fault data set of the semi-physical entity obtained in step S3, and the trained diagnostic model is actually applied in the physical entity of the complex equipment. The training process and diagnostic results are visualized in the virtual entity of the complex equipment.

Images