A marine hydraulic system fault diagnosis system and device

Abstract

The present application belongs to the field of marine hydraulic system, and particularly relates to a marine hydraulic system fault diagnosis system and device, comprising the following steps: constructing a knowledge base based on a fault tree structure; obtaining fault phenomenon description keywords and performing a first matching with the knowledge base; based on the first matching result and the first reasoning fault diagnosis result, continuing to obtain fault phenomenon description keywords and performing a second matching with the knowledge base; repeating the steps until a component-level fault is matched, then outputting the fault diagnosis result and recording it into weight analysis. The present application can greatly improve the diagnosis efficiency and accuracy based on fault tree and fault probability matching and reasoning. The present application has high integration and is convenient to carry; non-invasive measuring elements are adopted, which can non-contact measure sound emission, vibration, rotating speed, temperature and other signals, and can intelligently judge the fault according to the signals. Moreover, the present application can reduce the dependence on related personnel and reduce the difficulty of marine hydraulic system fault diagnosis.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of marine hydraulic systems, and particularly relates to a marine hydraulic system fault diagnosis system and device. BACKGROUND

[0002] At present, the marine hydraulic system fault is checked by artificial inspection, which has many disadvantages, one of which is that the detection of fault causes needs to use tools, and different types of faults need different detection tools, which on the one hand increases the cost and on the other hand is not easy to carry, and is easy to be lost; the second is that the inspection, analysis and elimination of the fault reason depend on the knowledge and experience of the maintenance personnel, and such talents are not easy to find; the third is that the existing detection means usually uses intervention sensor elements, which requires the measured equipment itself to reserve a measurement interface, which is usually difficult to meet, and during detection, it usually needs to stop installation, disassemble the pipeline and valve, and the detection and diagnosis efficiency is low, and the safe and continuous operation of the equipment is affected; the fourth is that there is currently a lack of effective detection means for hydraulic control valve leakage detection, and the troubleshooting efficiency of such faults is low. SUMMARY

[0003] The marine hydraulic system fault diagnosis system and device provided by the application can effectively solve the problems in the background art.

[0004] The marine hydraulic system fault diagnosis system provided by the application comprises the following steps:

[0005] S1: constructing a knowledge base based on a fault tree structure; the fault tree structure is classified according to the fault types of the hydraulic system, and each type of fault is hierarchically refined until the element-level fault at the bottom layer;

[0006] S2: obtaining fault phenomenon description keywords and performing a first matching with the knowledge base, outputting a fault diagnosis result when the first matching reaches the element-level fault and recording into weight analysis; when the first matching does not reach the element-level fault, outputting a first reasoning fault diagnosis result according to the weight analysis, and if the first reasoning fault diagnosis result is the final output fault diagnosis result, ending and recording into the weight analysis, otherwise, entering step S3;

[0007] S3: based on the first matching result and the first reasoning fault diagnosis result, continuing to obtain fault phenomenon description keywords and performing a second matching with the knowledge base, outputting a fault diagnosis result when the second matching reaches the element-level fault and recording into weight analysis; when the second matching does not reach the element-level fault, outputting a second reasoning fault diagnosis result according to the weight analysis, and if the second reasoning fault diagnosis result is the final fault diagnosis result, ending and recording into the weight analysis, otherwise, entering step S4;

[0008] S4: Loop steps S2 and S3 until a component-level fault is matched, then output the fault diagnosis result and record it in the weight analysis.

[0009] As a further optimization of the present invention, in step S3 , the keywords for the description of the fault phenomenon are continuously acquired by adopting a closed acquisition method, that is, node options of the next level are provided for selection based on the level matched last time.

[0010] As a further optimization of the present invention, the knowledge content of the component-level fault in step S1 includes fault detection indicators and maintenance methods.

[0011] As a further optimization of the present invention, the hydraulic system failure types include five categories: insufficient system pressure, unstable flow, actuator inaction, slow action and abnormal vibration.

[0012] As a further optimization of the present invention, the knowledge base in step S1 constructs the fault tree by combining the production expression method and the framework expression method.

[0013] The present invention also provides a marine hydraulic system fault diagnosis device, comprising a fault information acquisition module, a signal processing module and a fault diagnosis module;

[0014] The fault information acquisition module detects or receives hydraulic system fault information and sends it to the information processing module; the signal processing module converts the received fault information into an electrical signal and sends it to the fault diagnosis module; the fault diagnosis module outputs based on the marine hydraulic system fault diagnosis system.

[0015] As a further optimization of the present invention, the fault information acquisition module adopts non-contact sensors, including an acoustic emission sensor, an acceleration sensor, a rotation speed sensor and a temperature sensor.

[0016] As a further optimization of the present invention, the signal processing module includes a signal conditioning circuit, an A / D conversion circuit and an embedded microprocessor.

[0017] As a further optimization of the present invention, the fault diagnosis module is a tablet computer.

[0018] As a further optimization of the present invention, it includes a diagnostic box; the fault information acquisition module, the signal processing module and the fault diagnosis module are all integrated in the diagnostic box.

[0019] The marine hydraulic system fault diagnosis system provided by the application can greatly improve the diagnosis efficiency and accuracy based on fault tree and fault probability matching and reasoning. The marine hydraulic system fault diagnosis device provided by the application has high integration and is convenient to carry. Non-intervention measurement elements are adopted to non-contact measure sound emission, vibration, rotating speed, temperature and other signals, and can intelligently judge the fault according to the signals. Moreover, the application can reduce the dependence on related personnel and reduce the difficulty of marine hydraulic system fault diagnosis. BRIEF DESCRIPTION OF DRAWINGS

[0020] Figure 1 is a system pressure deficiency or pressure fluctuation fault tree in example 1;

[0021] Figure 2 is a structural schematic diagram of example 2;

[0022] Figure 3 is an appearance diagram of the diagnosis box of example 2. DETAILED DESCRIPTION

[0023] Example 1

[0024] The example includes the following steps:

[0025] S1: Construct a knowledge base based on a fault tree structure. The fault tree structure is classified according to the fault types of the hydraulic system, and each type of fault is hierarchically refined until the element level fault at the bottom.

[0026] The example mainly comprehensively collates the fault information of the experts, enterprise experts, technical personnel and maintenance personnel on the fault phenomenon, fault cause analysis and fault solving method of the hydraulic fault, represents the information according to the knowledge representation method suitable for expression, and establishes a knowledge base.

[0027] The knowledge base is established based on the fault tree mode, and they are associated with each other. One fault phenomenon represents one node, which may be the conclusion of a high-level fault node or the leading point of a lower-level fault. Therefore, in order to better represent it and reflect its characteristics, when considering the knowledge representation method, the hydraulic system fault characteristics are combined, the production rule and the frame representation method are adopted, and they are fused according to a certain structure.

[0028] (1) Production representation method

[0029] Production representation can simulate the relationship between various knowledge of experts (fault cause, solution, etc.), and its expression property is P→Q, where P represents conditions such as fault phenomenon, premise, state, etc., and Q represents results such as conclusion, action, etc. That is, if P is satisfied, Q can be inferred, such as P (pump vibration), and then Q (pump suction, bearing damage).

[0030] Production representation is simple and clear, and is very consistent with the mode of human thinking, and knowledge representation is concise, accurate, and easy for reasoning, but it has poor ability to express knowledge hierarchy, and its explanation ability is also limited, and it can only express cause-and-effect relationship knowledge.

[0031] (2) Frame representation

[0032] Frame representation is based on frame theory, and its basic idea is that a frame is a structural mode stored in the brain memory, and when a person faces a new scene or a fundamental change occurs to an existing view of a problem, he generally selects a basic knowledge structure called a frame from his memory, and the frame is a knowledge empty frame in his previous memory, and its specific content can be modified according to actual conditions. Frame theory takes frame as a unit of knowledge representation, and a group of related frames are combined to form a frame system, and different frames in the system can share a sub-frame, and system behavior can be represented by changes between frames. The frame is a network composed of a plurality of nodes and relationships, and the frame is composed of a frame name and a group of slots (Slot), and each slot has corresponding information.

[0033] Frame representation has the advantages of being good at expressing structural knowledge, and frames can form hierarchical or more complex relationships, and can reflect human thinking activities when observing things, and its disadvantages are that it is not convenient to express process knowledge. Hydraulic system fault contains a large amount of structural information such as fault phenomenon, fault cause, fault solution, etc., and if frame is simply used for expression, the logical relationship between information is relatively complex, and at the same time, due to the huge amount of information, the efficiency of reasoning may be reduced in operation.

[0034] Therefore, the embodiment adopts the combination of production and frame to express, and provides a better construction environment.

[0035] The production-frame representation method is to record fault information into the frame, and embed production rules into the frame, and in system operation, the information and rules in the frame are integrated with each other, and can be called by each other.

[0036] In order to transform the extended fault tree knowledge into production rule knowledge, the extended fault tree needs to be transformed into a set of "minimal" fault trees, i.e. each fault tree contains only one logical gate (AND gate, OR gate), and each minimal fault tree can be converted into one or more production rules in the diagnostic knowledge base: an "AND gate" corresponds to one rule; an "OR gate" can be transformed into multiple rules equal to the number of input OR gate events.

[0037] The fault tree analysis method is a method of representing and analyzing information such as fault phenomena and fault causes in a tree structure. It analyzes and judges the top-level fault phenomenon, finds the intermediate-level causes of the fault occurrence, and connects them with logical relationships, and then analyzes layer by layer until the lowest-level causes are found, forming a tree diagram. The fault tree analysis method has the characteristics of clear structure hierarchy and clear cause-and-effect relationship.

[0038] The knowledge base of the present embodiment is established based on a fault tree. When building the knowledge base, all fault information obtained through the knowledge acquisition system is processed, which can make the fault diagnosis result more accurate. All fault information such as fault phenomena, fault causes, and fault solutions obtained through the knowledge acquisition system is established through the structure of the fault tree. The mutual relationship between the upper and lower nodes is analyzed, the minimal cut sets of the fault tree, i.e. the element-level fault sets at the bottom layer, are obtained, and their relationship is organized according to the structure of the relationship mode table to form the knowledge base. Then, the knowledge of the knowledge base is expressed in the most intuitive fault tree diagram, which can ensure the completeness and compatibility of the knowledge.

[0039] The typical system-level faults of the marine hydraulic system mainly include insufficient system pressure, unstable flow, non-action of the actuator, slow action, and abnormal vibration. Therefore, the hydraulic system faults set in the present embodiment mainly include five types of insufficient system pressure, unstable flow, non-action of the actuator, slow action, and abnormal vibration. Other types can also be added in other embodiments, depending on the specific situation.

[0040] The fault causes of the main elements and auxiliary elements of the hydraulic system are diversified due to the correlation between them. A certain fault phenomenon can be caused by several factors, and a certain fault cause can also cause multiple fault phenomena in other parts of the associated hydraulic system. For example, Figure 1 The simultaneous occurrence of X2 hydraulic oil pollution and X14 filter blockage is the reason for the serious blockage of the M8 filter, and X1 incorrect parameter setting and X2 hydraulic oil pollution will cause the M3 hydraulic valve pressure regulating error as long as one of them occurs. In addition, X2 hydraulic oil pollution can cause M3 hydraulic valve pressure regulating error or M8 filter blockage. Therefore, the relationship between them needs to be specifically represented in the form of a fault tree during the establishment of the knowledge base.

[0041] S2: Obtain the fault phenomenon description keywords and match them with the knowledge base. When a component-level fault is matched, output the fault diagnosis result. In this embodiment, the output fault diagnosis result includes the fault cause, fault detection indicators, and maintenance methods. The fault diagnosis result is verified by the fault detection indicators, and the fault is repaired in a timely manner. At the same time of outputting the fault diagnosis result, the fault diagnosis result is recorded and analyzed to obtain the weight.

[0042] When a component-level fault is not matched, the one-time reasoning fault diagnosis result is output according to the weight analysis. If the one-time reasoning fault diagnosis result is the final output fault diagnosis result, the process ends, and the fault diagnosis result is recorded and analyzed in the weight analysis. Otherwise, step S3 is entered.

[0043] S3: Continue to obtain the fault phenomenon description keywords based on the one-time matching result and the one-time reasoning fault diagnosis result, and perform secondary matching with the knowledge base. Here, based on the one-time matching result and the one-time reasoning result means that all paths from the bottom layer to the highest layer containing component-level faults matched and reasoned out previously are excluded in the fault tree. Figure 1 For example, when the X2 hydraulic oil pollution is excluded, all the upper layers and paths related to X2 are also excluded. For example, M8 filter blockage is excluded. Similarly, if there is another upper layer related to M8, it is also excluded.

[0044] The system provided in this embodiment excludes paths from bottom to top based on matching results and reasoning results to continuously narrow down the results until the final result is found. Since fault detection of marine hydraulic systems is based on the most basic component-level fault, only component-level faults can be accurately detected by sensors. Therefore, whether matching or reasoning starts from component-level faults, which is the most direct and easiest method to detect and verify, and has the highest accuracy.

[0045] When a component-level fault is matched in the secondary matching, the fault diagnosis result is output and recorded in the weight analysis. When a component-level fault is not matched in the secondary matching, the secondary reasoning fault diagnosis result is output according to the weight analysis. If the secondary reasoning fault diagnosis result is the final fault diagnosis result, the process ends and is recorded in the weight analysis. Otherwise, step S4 is entered.

[0046] S4: Repeat steps S2 and S3 until a component-level fault is matched, and output the fault diagnosis result and record it in the weight analysis.

[0047] Since one fault phenomenon may correspond to multiple fault components or causes, a preliminary judgment is made according to the weight of each fault component during initial use. As the diagnosis data accumulates, the weight of the fault component is continuously updated to better fit the actual system.

[0048] The step S continues to acquire the fault phenomenon description keywords, which can be acquired in an open manner or in a closed manner.

[0049] The open manner refers to extracting keywords from inputted arbitrary information to acquire information.

[0050] The closed manner refers to providing options of the next level based on the last matched level. Still taking the example in the above, when the fault cause is matched to be M1 system pressure deficiency, M3 hydraulic valve pressure adjustment error, M4 pipe joint sealing damage oil leakage, M5 mailbox oil shortage and M hydraulic pump fault are provided for selection. Figure 1

[0051] Embodiment 2

[0052] The embodiment provides a marine hydraulic system fault diagnosis device, which comprises a fault information acquisition module, a signal processing module and a fault diagnosis module.

[0053] As shown in FIG. 1, the fault information acquisition module receives the fault information of the hydraulic system and sends the fault information to the information processing module. The fault diagnosis module can be externally connected with input devices such as a mouse, a keyboard and a touch display, and the fault information provided by a person can be acquired through the input devices. Figure 2

[0054] The fault information acquisition module can also adopt non-contact sensors, including detection devices such as acoustic emission sensors, acceleration sensors, speed sensors and temperature sensors. The channel number of the sensor signal acquisition in the embodiment is 5, and the temperature, speed, acoustic emission signal and vibration acceleration data of the main components such as the hydraulic pump, the vane pump and the electromagnetic valve in the whole ship hydraulic system can be acquired in real time. The non-contact sensors are adopted in the embodiment to realize non-destructive testing of elements, and the acoustic emission signal acquisition can process and analyze the collected acoustic emission signal of the leakage in the spool, provide signal time domain waveform, amplitude spectrum, power spectrum, energy proportion analysis and can quantitatively evaluate the internal leakage state.

[0055] The fault information acquisition module in the embodiment comprises detection devices and externally connected devices, and the fault information is comprehensively acquired in combination of the two manners.

[0056] The signal processing module converts the received fault information into an electrical signal and sends the electrical signal to the fault diagnosis module. Further, the signal processing module comprises a signal conditioning circuit, an A / D conversion circuit and an embedded microprocessor.

[0057] ​​The signal conditioning circuit mainly consists of signal amplification, rectification filtering and level conversion circuits, and can realize the collection of voltage type sensor signals and the collection of current type sensor signals. The signals after conditioning shall have certain quality, so as to ensure the accuracy and reliability of the analog signals input to the A / D conversion circuit. The A / D conversion circuit can convert the analog signals of the sensor into digital signals.

[0058] The embedded microprocessor as the control core of the lower computer needs to have the functions of external storage, USB interface, network interface, RS485 interface (reserved) and power module. The network interface is responsible for communication with the lower computer, and completes the uploading and downloading of monitoring data during state monitoring. The RS-485 interface is a reserved interface, responsible for communication with the ship hydraulic system pump station PLC, and realizes the acquisition of the internal system state parameters of the PLC, which can enrich the data quantity for fault diagnosis.

[0059] The fault diagnosis module is based on the marine hydraulic system fault diagnosis system in example 1. The fault diagnosis module uses a tablet computer. The liquid crystal touch screen on the tablet computer can display the real-time collected system state parameters and monitoring data. Due to the large number of components of the marine hydraulic system, but the size of the liquid crystal touch screen is limited, it is impossible to display all components at one time. In order to facilitate the selection of components, the components of the ship hydraulic system can be divided into several modules according to function or region, and the corresponding function buttons are configured to realize the processing, display and storage of data, and the fault diagnosis of single components.

[0060] As shown in Figure 3 Further, a diagnosis box is included. The fault diagnosis module, the signal processing module and the fault diagnosis module are integrated in the diagnosis box.

[0061] All electronic components are uniformly installed in the sealed diagnosis box. The diagnosis box adopts certain damping, sealing, corrosion prevention and anti-interference measures, which can ensure the reliable work of all electronic components in the ship vibration, oil pollution, humidity, multiple salt and complex electromagnetic environment, and ensure the reliability and stability of the fault diagnosis box in the field use.

[0062] Finally, it should be noted that the above examples are only used to illustrate the technical solutions of the present application, and are not limited to the scope of protection of the present application. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced by equivalents without departing from the essence and scope of the technical solutions of the present application.

Claims

1. A marine hydraulic system fault diagnosis system, characterized in that: The following steps are involved: S1: Build a knowledge base based on the fault tree structure; The fault tree structure is based on the classification of hydraulic system fault types, and each type of fault is refined hierarchically until the lowest component-level fault; S2: Obtain the fault phenomenon description keywords and perform a match with the knowledge base. If a component-level fault is matched, the fault diagnosis result is output and recorded in the weight analysis. If a component-level fault is not matched, an inference fault diagnosis result is output based on the weight analysis. If the inference fault diagnosis result is the final output fault diagnosis result, the process ends and is recorded in the weight analysis. Otherwise, the process proceeds to step S3. S3: Based on the primary matching result and the primary inference fault diagnosis result, the fault phenomenon description keywords are continuously obtained and matched against the knowledge base for a secondary match. If the secondary match finds a component-level fault, the fault diagnosis result is output and recorded in the weight analysis. If the secondary match does not find a component-level fault, the secondary inference fault diagnosis result is output based on the weight analysis. If the secondary inference fault diagnosis result is the final fault diagnosis result, the process ends and is recorded in the weight analysis. Otherwise, the process proceeds to step S4. S4: loop through steps S2 and S3 until a component-level fault is found, output the fault diagnosis result, and record it in the weight analysis; In step S3, the keywords describing the fault phenomenon are obtained in a closed manner, that is, node options of the next level are provided for selection based on the level matched last time; The knowledge content of component-level faults in step S1 includes fault detection indicators and repair methods; Hydraulic system failure types include insufficient system pressure, unstable flow, actuator failure, slow movement and abnormal vibration.

2. A marine hydraulic system fault diagnosis system according to claim 1, characterized in that: The knowledge base in step S1 constructs a fault tree by combining production representation and framework representation.

3. A marine hydraulic system fault diagnosis device, characterized in that: It includes fault information acquisition module, signal processing module and fault diagnosis module; The fault information acquisition module detects or receives hydraulic system fault information and sends it to the information processing module; the signal processing module converts the received fault information into an electrical signal and sends it to the fault diagnosis module; the fault diagnosis module outputs based on the marine hydraulic system fault diagnosis system according to claim 1.

4. A marine hydraulic system fault diagnosis device according to claim 3, characterized in that: The fault information acquisition module adopts non-contact sensors, including acoustic emission sensors, acceleration sensors, speed sensors and temperature sensors.

5. A marine hydraulic system fault diagnosis device according to claim 3, characterized in that: The signal processing module includes a signal conditioning circuit, an A / D conversion circuit and an embedded microprocessor.

6. A marine hydraulic system fault diagnosis device according to claim 3, characterized in that: The fault diagnosis module is a tablet computer.

7. A marine hydraulic system fault diagnosis device according to claim 3, characterized in that: It comprises a diagnosis box; a fault information acquisition module, a signal processing module and a fault diagnosis module are all integrated in the diagnosis box.

Images