CN113791923B - A method for fault detection and fault handling of an underwater vehicle - Google Patents

Abstract

The invention belongs to the technical field of fault detection and fault processing of an underwater vehicle control system, and particularly relates to a fault detection and fault processing method of the underwater vehicle control system, wherein the control system comprises the following steps: a master control node and a plurality of controlled nodes; comprising the following steps: the master control node sends a broadcasting instruction of a period T to the controlled node through a digital bus; the main control node judges and generates a corresponding fault record according to the response condition and response content of the controlled node, and stores the fault record as record-level fault data to finish fault detection of the controlled node; the main control node maps the record level fault data to the corresponding processing level fault data in real time according to an externally issued debugging and real-time navigation conversion instruction, and the main control node adopts a fault active processing method to process the fault according to the content of the processing level fault data; the controlled node detects the communication state between the controlled node and the main control node, judges the communication fault, and carries out the passive fault processing by the controlled node.

Description

A method for fault detection and fault handling of an underwater vehicle

technical field

The invention belongs to the technical field of fault detection and fault processing of an underwater vehicle control system, and in particular relates to a fault detection and fault processing method for an underwater vehicle control system.

Background technique

With the increasing maturity and widespread use of digital and modular design technologies, the internal control nodes of underwater vehicles such as propulsion motor drivers, rudder motor drivers, battery management system (Battery Management System, BMS) components, inertial navigation sensors, pressure gauges, etc. The design of the tightly coupled mode is no longer used among the underwater vehicles, but the data exchange is carried out through the digital bus. The establishment and use of the digital bus makes the function design of each controlled node in the underwater vehicle more independent, thereby reducing the design and Manufacturing difficulty reduces maintenance and upgrade costs.

Motion control is one of the core functions of underwater vehicles. During the digital communication process of the control system, the main control node set on the main CPU (Central Processing Unit, referred to as the central processing unit) of the underwater vehicle control system controls each controlled node. The data acquisition and command control work of the computer is realized through the digital bus (that is, digital communication). In order to ensure the reliable operation of the underwater vehicle, it is necessary to monitor the faults generated by the control system in real time, and perform corresponding fault handling when the fault of the corresponding fault type occurs.

However, the existing fault handling methods are single and limited, not universal, and greatly reduce work efficiency, and cannot perform fault detection and fault processing on multiple controlled nodes at the same time.

Contents of the invention

The invention proposes a widely applicable and easily expandable fault detection and processing method. The control system includes: a master control node and a plurality of controlled nodes; the master control node and the controlled nodes perform master-slave communication through a digital bus; The internal communication of the control system and each controlled node perform fault detection and fault processing at the same time;

Aiming at a controlled node set inside the underwater vehicle, the main control node communicates with the controlled node through a digital bus, and adopts this method to perform fault detection and fault processing on the controlled node;

The method includes:

The master control node sends a broadcast command of period T to the controlled node through the digital bus;

The main control node judges and generates corresponding fault records according to the response status and response content of the controlled node, and stores them as record-level fault data to complete the fault detection of the controlled node;

The main control node maps the record-level fault data to the corresponding processing-level fault data in real time according to the external commissioning and actual flight instructions, and the main control node adopts the fault active processing method to handle the fault according to the content of the processing-level fault data ;

The controlled node detects its communication status with the main control node, judges the communication failure, and the controlled node handles the failure passively.

As one of the improvements of the above technical solution, before implementing the method, it also includes: defining 32-bit CER variable, 32-bit UER variable, 32-bit DER variable, 32-bit CED variable, 32-bit UED variable and 32-bit DED variable;

The 32-bit CER variable is used as a buffer for record-level communication failures;

The 32-bit UER variable is used as a cache for record-level data update exception faults;

The 32-bit DER variable is used as a cache for record-level data exception faults;

The 32-bit CED variable is used as a buffer for processing-level communication faults, corresponding to the processing-level mapping of CER variables;

The 32-bit UED variable is used as a cache for processing-level data update exception faults, corresponding to the processing-level mapping of UER variables;

The 32-bit DED variable is used as a buffer for processing-level data exception faults, corresponding to the processing-level mapping of DER variables;

The data bits in the 32-bit CER variable are mapped one-to-one with the data bits in the 32-bit CED variable;

The data bits in the 32-bit UER variable are mapped one-to-one with the data bits in the 32-bit UED variable;

The data bits in the 32-bit DER variable are mapped one-to-one with the data bits in the 32-bit DED variable;

The various variables mentioned above are all reset to 0 when they are initialized.

As one of the improvements to the above technical solution, the master control node judges and generates a corresponding fault record according to the response and response content of the controlled node, and stores it as record-level fault data to complete the The fault detection of the control node; the specific process is:

The main control node judges and generates a communication fault record according to the response of the controlled node, and stores it as record-level communication fault data, and determines the fault type as a communication fault;

According to the response content of the controlled node, the master control node judges and generates a functional fault record, stores it as record-level functional fault data, and determines the fault type as a functional fault.

As one of the improvements to the above technical solution, the master control node judges and generates a communication fault record according to the response of the controlled node, stores it as record-level communication fault data, and determines the fault type as a communication fault; The specific process is:

The master control node sends a broadcast command with a period of T to the controlled node, and the controlled node responds;

If the response of the controlled node is detected within the period T, the master control node controls the value of the communication abnormality counter corresponding to the controlled node to be cleared to 0;

If the response of the controlled node is not detected within the period T, the master control node controls the value of the communication abnormality counter corresponding to the controlled node to be increased by 1;

Then count the value of the abnormal communication counter, record it as a statistical value, and compare the statistical value with a preset threshold;

If the statistical value is greater than or equal to the preset threshold, set the bit value of the data bit corresponding to the controlled node in the 32-bit CER variable to 1 to generate a communication fault record and update the record-level communication fault data , store it, and determine that the fault type is a communication fault;

If the statistical value is less than the preset threshold, no processing will be done.

As one of the improvements of the above technical solution, the master control node judges and generates a functional fault record according to the response content of the controlled node, stores it as record-level functional fault data, and determines the fault type as a functional fault; The specific process is:

The master control node sends a broadcast command with a period of T to the controlled node, and the controlled node responds;

If the response of the controlled node is detected in the period T, record the content of the response of the controlled node in the current period, as the functional data of the response, record it as the functional data of the current period, and judge the difference between the functional data of the current period and the previous period Whether the functional data of is the same;

If the function data of the current cycle is different from the function data of the previous cycle, the master control node controls the value of the data refresh exception counter corresponding to the controlled node to be cleared to 0;

If the function data of the current cycle is the same as the function data of the previous cycle, the master control node controls the value of the data refresh exception counter corresponding to the controlled node to add 1;

Then count the value of the data refresh abnormal counter, record it as the refresh statistical value, and compare the refresh statistical value with the preset refresh threshold;

If the refresh statistics value is less than the preset refresh threshold, no processing will be done;

If the refresh statistic value is greater than or equal to the preset refresh threshold, the bit value of the data bit corresponding to the controlled node in the 32-bit UER variable is set to 1, so as to generate a data refresh abnormal fault record and update the record Level data refreshes the fault data, stores it, and determines that the fault type is abnormal data refresh in functional faults;

At the same time, it is judged whether the functional data of the current period exceeds the preset boundary range;

If the functional data of the current cycle exceeds the preset boundary range, set the bit value of the data bit corresponding to the controlled node in the 32-bit DER variable to 1 to generate a data abnormal fault record and update the record-level data Abnormal fault data is stored, and the fault type is determined to be abnormal data in functional faults;

If the functional data of the current cycle does not exceed the preset boundary range, no processing will be done.

As one of the improvements of the above-mentioned technical solution, the main control node maps the record-level fault data to the corresponding processing-level fault data in real time according to the instruction issued by the outside for commissioning to actual flight, and the main control node maps the fault data at the processing level according to the content, use the active fault handling method to handle the fault; the specific process is as follows:

When the command to transfer from commissioning to actual flight is issued externally, the following operations are performed in each subsequent cycle:

Copy and update the record-level communication fault data to the processing-level communication fault data, specifically, copy and update the data bit of the CER variable to the data bit of the CED variable;

Update the record-level data refresh fault data copy to the processing-level data refresh fault data, specifically, update the data bit copy of the UER variable to the data bit of the UED variable;

Copy and update record-level data abnormal fault data to processing-level data abnormal fault data, specifically, update and copy the data bits of the DER variable to the data bits of the DED variable;

If any position of the data bit of the CED variable, the data bit of the UED variable or the data bit of the DED variable is greater than 0, the fault processing method is adopted to handle the fault actively;

If all positions of the data bits of the CED variable, the data bits of the UED variable, or the data bits of the DED variable are all equal to 0, then do nothing;

As one of the improvements to the above technical solution, the method further includes: each controlled node detects the communication status between itself and the main control node, and judges according to the detected reception of the broadcast instruction periodically sent by the main control node And generate communication fault records, store them on the controlled node, and determine the type of fault as communication fault; the controlled node adopts the method of fault passive processing according to its own function type to handle the fault.

As one of the improvements of the above technical solution, each controlled node detects the communication status between itself and the main control node, and judges and generates a communication fault according to the detected reception of the broadcast instruction periodically sent by the main control node Record, store on the controlled node, and determine the type of fault as a communication fault; the controlled node adopts a passive fault handling method according to its own function type to handle the fault; the specific process is:

The controlled node internally sets a timer with the same period T as the broadcast command of the master control node, and a passive communication exception counter;

If the controlled node detects the master control node broadcast command within the timer period T, the value of the passive communication exception counter set inside the controlled node is cleared to 0;

If the controlled node does not detect the broadcast command of the master control node within the timer period T, the value of the passive communication exception counter set inside the controlled node is increased by 1;

Then count the value of the abnormal communication counter, record it as a statistical value, and compare the statistical value with a preset threshold;

If the statistical value is greater than or equal to the preset threshold, a communication fault record is generated and stored on the controlled node, and the fault type is determined to be a communication fault; the controlled node adopts a passive fault handling method according to its own function type to carry out Troubleshooting

If the statistical value is less than the preset threshold, no processing will be done.

The beneficial effect of the present invention compared with prior art is:

The method of the present invention can detect and process the faults of each controlled node inside the underwater vehicle in real time, has universal applicability, and can be flexibly expanded with the expansion of the function of the vehicle, is convenient, simple and easy to implement, and greatly improves Improve work efficiency; at the same time, because the method of the present invention can also be applicable to any electronic equipment that has digital bus to carry out master-slave communication, has universal applicability.

Description of drawings

Fig. 1 is a structural schematic diagram of fault types and fault processing involved in an underwater vehicle fault detection and fault processing method of the present invention;

Fig. 2 is the structural diagram of the underwater vehicle control system applicable to the present invention;

Fig. 3 is an internal structure diagram of the underwater vehicle in a method for detecting and handling the failure of the underwater vehicle according to the present invention;

Fig. 4 is a detection flow chart of the fault detection in the fault detection and fault processing method of the underwater vehicle according to the present invention.

Detailed ways

The present invention will be further described in conjunction with accompanying drawing and example now.

As shown in Figures 1 and 2, the present invention also provides a fault detection and fault handling method for an underwater vehicle,

The control system includes: a main control node and multiple controlled nodes; the main control node and the controlled nodes perform master-slave communication through a digital bus; the internal communication of the control system and each controlled node simultaneously perform fault detection and fault processing;

Among them, the main control node and the controlled node correspond to each controlled node set inside the underwater vehicle; among them, the controlled node with the main control function is the main control node, and the controlled node controlled by the main control node is the controlled node .

Aiming at a controlled node set inside the underwater vehicle, the main control node communicates with the controlled node through a digital bus, and adopts this method to perform fault detection and fault processing on the controlled node;

The method includes:

Step 1) The master control node sends a broadcast command of period T to the controlled node through the digital bus;

Step 2) The main control node judges and generates a corresponding fault record according to the response situation and response content of the controlled node, and stores it as record-level fault data to complete the fault detection of the controlled node;

Specifically, the master control node judges and generates a communication failure record according to the response of the controlled node, stores it as record-level communication failure data, and determines the failure type as a communication failure;

Specifically, the master control node sends a broadcast instruction with a period of T to the controlled node, and the controlled node responds;

If the response of the controlled node is detected within the period T, the master control node controls the value of the communication abnormality counter corresponding to the controlled node to be cleared to 0;

If the response of the controlled node is not detected within the period T, the master control node controls the value of the communication abnormality counter corresponding to the controlled node to increase by 1; wherein, the communication abnormality counter is set on the master control node;

Then count the value of the abnormal communication counter, record it as a statistical value, and compare the statistical value with a preset threshold;

If the statistical value is greater than or equal to the preset threshold, set the bit value of the data bit corresponding to the controlled node in the 32-bit CER variable to 1 to generate a communication fault record and update the record-level communication fault data , store, and determine that the fault type is a communication fault; wherein, the stored communication fault record is stored on the master control node;

If the statistical value is less than the preset threshold, no processing will be done.

In this embodiment, the communication anomaly counter is CntCommM; the communication anomaly counter is initialized and reset to 0.

According to the response content of the controlled node, the master control node judges and generates a functional fault record, stores it as record-level functional fault data, and determines the fault type as a functional fault.

Specifically, the master control node sends a broadcast instruction with a period of T to the controlled node, and the controlled node responds;

If the response of the controlled node is detected in the period T, record the content of the response of the controlled node in the current period, as the functional data of the response, record it as the functional data of the current period, and judge the difference between the functional data of the current period and the previous period Whether the functional data of is the same;

If the function data of the current cycle is different from the function data of the previous cycle, the master control node controls the value of the data refresh exception counter corresponding to the controlled node to be cleared to 0;

If the functional data of the current cycle is the same as the functional data of the previous cycle, the master control node controls and adds 1 to the value of the data refresh abnormal counter corresponding to the controlled node; wherein, the data refresh abnormal counter is set on the master control node;

Then count the value of the data refresh abnormal counter, record it as the refresh statistical value, and compare the refresh statistical value with the preset refresh threshold;

If the refresh statistics value is less than the preset refresh threshold, no processing will be done;

If the refresh statistic value is greater than or equal to the preset refresh threshold, the bit value of the data bit corresponding to the controlled node in the 32-bit UER variable is set to 1, so as to generate a data refresh abnormal fault record and update the record Level data refreshes fault data, stores, and determines that the fault type is abnormal data refresh in functional faults; wherein, the stored data refresh fault data is stored on the main control node;

At the same time, it is judged whether the functional data of the current period exceeds the preset boundary range;

If the functional data of the current cycle exceeds the preset boundary range, set the bit value of the data bit corresponding to the controlled node in the 32-bit DER variable to 1 to generate a data abnormal fault record and update the record-level data The abnormal fault data is stored, and the fault type is determined to be data abnormal in functional fault; wherein, the stored data abnormal fault data is stored on the main control node;

If the functional data of the current cycle does not exceed the preset boundary range, no processing will be done.

Wherein, in this embodiment, the data refresh abnormal counter is CntUpdateM, which is reset to 0 when it is initialized; it is reset to 0 when the function data is initialized;

Step 3) The main control node maps the record-level fault data to the corresponding processing-level fault data in real time according to the externally issued instruction for commissioning to actual flight, and the main control node adopts the fault active processing method according to the content of the processing-level fault data, Carry out fault handling; specifically, when the external commissioning command is sent to actual flight, the following operations are performed in each subsequent cycle:

Copy and update the record-level communication fault data to the processing-level communication fault data, specifically, copy and update the data bit of the CER variable to the data bit of the CED variable;

Update the record-level data refresh fault data copy to the processing-level data refresh fault data, specifically, update the data bit copy of the UER variable to the data bit of the UED variable;

Copy and update record-level data abnormal fault data to processing-level data abnormal fault data, specifically, update and copy the data bits of the DER variable to the data bits of the DED variable;

If any position of the data bit of the CED variable, the data bit of the UED variable or the data bit of the DED variable is greater than 0, the fault processing method is adopted to handle the fault actively;

If all positions of the data bits of the CED variable, the data bits of the UED variable, or the data bits of the DED variable are all equal to 0, then do nothing;

Step 4) The controlled node detects its communication status with the main control node, judges the communication failure, and the controlled node performs passive processing of the failure.

The method also includes: step 5) each controlled node detects the communication state between it and the main control node, and judges and generates a communication failure record according to the detected reception of the broadcast instruction periodically sent by the main control node, Storage is carried out on the controlled node, and the fault type is determined to be a communication fault; the controlled node adopts a fault passive processing method to handle the fault according to its own function type.

Among them, this step is to consider that when the communication between the master control node and the controlled node is disconnected, the master control node cannot obtain the response of the controlled node. , to judge the communication fault, and the controlled node completes the fault processing by itself. This process is not controlled by the master control node, so it is called fault passive processing.

Specifically, the controlled node internally sets a timer with the same period T as the broadcast command of the master control node, and a passive communication exception counter;

If the controlled node detects the master control node broadcast command within the timer period T, the value of the passive communication exception counter set inside the controlled node is cleared to 0;

If the controlled node does not detect the broadcast command of the master control node within the timer period T, the value of the passive communication exception counter set inside the controlled node is increased by 1;

Then count the value of the abnormal communication counter, record it as a statistical value, and compare the statistical value with a preset threshold;

If the statistical value is greater than or equal to the preset threshold, a communication fault record is generated and stored on the controlled node, and the fault type is determined to be a communication fault; the controlled node adopts a passive fault handling method according to its own function type to carry out Troubleshooting

If the statistical value is less than the preset threshold, no processing will be done.

Wherein, before implementing the method, it also includes: defining 32-bit CER variables, 32-bit UER variables, 32-bit DER variables, 32-bit CED variables, 32-bit UED variables and 32-bit DED variables;

The 32-bit CER variable (CommErrRecord variable) is used as a buffer for record-level communication failures;

The 32-bit UER variable (the variable of UpdateErrRecord) is used as a cache for record-level data update exception faults;

The 32-bit DER variable (DataErrRecord variable) is used as a cache for record-level data exception faults;

32-bit CED variable (CommErrDeal variable), corresponding to the processing level mapping of CER variable;

The 32-bit UED variable (UpdateErrDeal variable) corresponds to the processing-level mapping of the UER variable;

32-bit DED variable (DataErrDeal variable), corresponding to the processing level mapping of DER variable;

The data bits in the 32-bit CER variable are mapped one-to-one with the data bits in the 32-bit CED variable;

The data bits in the 32-bit UER variable are mapped one-to-one with the data bits in the 32-bit UED variable;

The data bits in the 32-bit DER variable are mapped one-to-one with the data bits in the 32-bit DED variable;

The various variables mentioned above are all reset to 0 when they are initialized.

Among them, the determination of the type of communication failure does not need to detect data, but only needs to detect whether the controlled node responds; the type of functional failure needs to determine the specific value. If there is no number, it has been judged as a communication failure.

Adopt above-mentioned step 1)-step 4) method simultaneously to each controlled node that underwater vehicle interior is provided with, can carry out fault detection to each controlled node simultaneously, have improved work efficiency greatly; Simultaneously, because the method of the present invention can It is applicable to any electronic equipment with digital bus for master-slave communication, and has universal applicability.

In the method of the present invention, controlled nodes with different functions can be added as required, and fault detection and fault processing can be performed on each controlled node inside the underwater vehicle at the same time, thereby greatly improving work efficiency.

Example 1.

In the system of the present invention, as shown in Figure 1, two types of faults can be detected by the method of the present invention: communication faults and functional faults. In the fault types, the generation of communication faults is based on periodic communication. The specific detection method is , by sending the broadcast command in the cycle T, each controlled node in the underwater vehicle responds to the function data of the corresponding controlled node after receiving the broadcast command, records and counts the responses of each controlled node every cycle, if If no response from the corresponding controlled node is received for N consecutive cycles, it is considered that the controlled node has a communication failure; similarly, if the controlled node does not receive the broadcast command from the master control node for N consecutive cycles, it is considered that the controlled node The control node has a communication failure.

Among the fault types, functional faults include data update anomalies and data anomalies. The abnormal data update is generated by monitoring the functional data of each controlled node. Specifically, if the functional data of the corresponding controlled node of the master control node does not change for M consecutive cycles, it is considered that the controlled node has an abnormal data update. If If the output data of the detected controlled node has fast-changing characteristics, the value of M should be set smaller. If the output data of the detected controlled node has slow-changing characteristics, the value of M should be set larger. If the controlled node The output changes very slowly, and data refresh detection may not be performed; among them, M value is the preset refresh threshold; data abnormality is generated by monitoring whether the functional data of each controlled node exceeds the preset boundary range, if any Once the functional data of the controlled node exceeds the preset boundary range, it is considered that the controlled node has a data anomaly.

In fault handling, fault records adopt a hierarchical record structure. The first level is the record level, and the second level is the processing level. These two levels have the same data structure type. Among them, each data bit in the record-level structure corresponds to a fault source, the bit value of the data bit is 1, indicating that the fault of the controlled node corresponding to the data bit is detected, and the bit value of the data bit is 0, indicating that the corresponding fault is not detected , all bits are initialized to 0 during initialization; after certain conditions are met, for example, the underwater vehicle changes from debugging mode to actual flight mode, and the data of the record level is periodically copied and updated to the data structure of the processing level. At this time, the data of the processing level is Mapping of record-level data.

During fault handling, the processing-level data is monitored in real time, and when it is detected that there is data greater than 0 in the processing-level data, the fault is actively handled.

In fault handling, when each controlled node detects a communication fault during the communication fault detection process or detects data refresh abnormality or data abnormality in the functional fault detection process during the functional fault detection process, it executes the predefined fault processing of the controlled node Action, the predefined fault handling mechanism of the controlled node is not controlled by the master control node, and is a passive processing of faults.

Example 2.

As shown in Figure 3, the underwater vehicle adopted in this embodiment is a basic electrodynamic underwater vehicle, and the main control CPU corresponding to the main control node communicates with the propulsion motor driver (hereinafter referred to as the propeller), Steering motor driver (hereinafter referred to as steering gear), battery BMS component (hereinafter referred to as battery), inertial navigation sensor (hereinafter referred to as inertial navigation), pressure sensor (hereinafter referred to as pressure gauge) communication. Wherein, thruster, steering gear, battery, inertial navigation and pressure gauge are the various components that underwater vehicle interior is provided with, and each component is as a corresponding controlled node in this embodiment; Main control CPU and each controlled node Carry out master-slave communication through the digital bus; perform fault detection and fault processing on the internal communication of the control system and each controlled node at the same time;

Before implementing this method, define a 32-bit variable named CommErrRecord as a buffer for record-level communication failures, where bit0 corresponds to the thruster, bit1 corresponds to the steering gear, bit2 corresponds to the battery, bit3 corresponds to the inertial navigation, and bit4 corresponds to the pressure gauge.

Define a 32-bit variable named UpdateErrRecord as a buffer for record-level data update abnormal faults, where bit0 corresponds to the thruster, bit1 corresponds to the steering gear, bit2 corresponds to the battery, bit3 corresponds to the inertial navigation, and bit4 corresponds to the pressure gauge.

Define a 32-bit variable named DataErrRecord as a buffer for record-level data abnormal faults, where bit0 corresponds to the thruster, bit1 corresponds to the steering gear, bit2 corresponds to the battery, bit3 corresponds to the inertial navigation, and bit4 corresponds to the pressure gauge.

Define the 32-bit CommErrDeal variable, which corresponds to the processing-level mapping of the CommErrRecord variable; define the 32-bit UpdateErrDeal variable, which corresponds to the processing-level mapping of the UpdateErrRecord variable; define the 32-bit DataErrDeal variable, which corresponds to the processing-level mapping of the DataErrRecord variable.

All the above variables are reset to 0 when the system is initialized.

The master control node sends a broadcast command in a period of 20 milliseconds, and each corresponding controlled node responds to the function data immediately after receiving the broadcast command, and the master control node sets a corresponding communication exception counter for each controlled node. At the end of the period, the response status of each controlled node is detected. If the response of the controlled node is not received in this period, the corresponding abnormal communication counter of the controlled node is increased by 1. If the abnormal communication counter is greater than or equal to 10 (predefined N=10, which is the preset threshold), then the value of the corresponding data bit in the CommErrRecord is set to 1, and if the response from the controlled node is received in this cycle, the corresponding communication abnormality counter is cleared to 0. For example, the communication exception counter corresponding to the current propeller is 9. If no response from the propeller is received in this cycle, the propeller communication exception counter is incremented by 1, which is 10 at this time, and it is judged that the propeller communication abnormal counter is greater than or equal to 10. Then the bit0 of CommErrRecord is set to 1; if the thruster’s response is received in this cycle, the thruster’s communication exception counter is cleared to 0.

As above, the master control node sends a broadcast command in a period of 20 milliseconds, and each corresponding controlled node responds to the functional data immediately after receiving the broadcast command, and a corresponding data refresh exception counter is set on each controlled node. At the end, the response content of each controlled node is detected. As the functional data of the response, if the functional data of this cycle has not changed compared with the functional data of the controlled node in the previous cycle, the corresponding data refresh abnormal counter will be incremented by 1. If this When the data refresh abnormal counter is greater than or equal to 15 (M=15 in the pre-defined), then the position value of the corresponding data bit in the UpdateErrRecord is set to 1, if the function data changes, then the corresponding data refresh abnormal counter is cleared to 0. For example, the data refresh exception counter corresponding to the current pressure gauge is 14. If the pressure gauge data in this cycle is equal to the pressure gauge data in the previous cycle, the pressure gauge data refresh exception counter is incremented by 1, and it is 15 at this time. It is judged that the pressure gauge data refresh exception counter is greater than If it is equal to 15, bit4 of UpdateErrRecord is set to 1; if the pressure gauge data in this cycle is not equal to the pressure gauge data in the previous cycle, the pressure gauge data refresh exception counter is cleared to 0.

At the end of any one cycle, if it is detected that the functional data of a controlled node exceeds the theoretical value range under its working conditions (that is, the preset boundary range), it is considered that the sensor is faulty, and the corresponding position in DataErrRecord is set to 1. For example, the dynamic range of the pressure gauge is 0 to 1000 meters, but the maximum working depth of the underwater vehicle is designed to be 100 meters. If the underwater vehicle is in normal operation, the pressure gauge has any sensor output that is obviously greater than 100 meters (such as 200 meters) , that is, it is judged that the pressure gauge is faulty, and at this time, bit4 of DataErrRecord is set to 1.

The detection method of the fault type is shown in Figure 4 below. The dotted line in the figure indicates the process before or after the fault detection is omitted, and the figure only shows the fault detection process of a single controlled node, and the remaining multiple controlled nodes The detection can be performed sequentially with reference to the detection flow shown in FIG. 4 .

The fault record level data is updated in real time every cycle, reflecting the real-time detection results; the processing level data is used to trigger the fault processing process. Since the underwater vehicle may have incomplete control nodes in the debugging state, it is hoped that Normal debugging operation instead of triggering exception handling. Therefore, only after an external command such as "debugging to actual flight" is issued, it is necessary to copy and update the record-level fault data to the processing-level fault data. The actual meaning of the "debugging to actual flight" command It becomes the fault processing enabling command. Once the "commissioning to actual flight" command is issued, the record-level data needs to be copied and updated to the processing-level data structure every cycle.

If the main control node detects that there is a non-zero value in the fault data at the processing level, it needs to handle the fault. For example, for an underwater vehicle with positive buoyancy, it needs to stop to float the product; for an underwater vehicle with negative buoyancy, it needs to dump the load or start the buoyancy device to float the product. This process is an active control and an active fault handling mechanism.

Each controlled node detects a fault and handles it passively. The failure of each controlled node is divided into two categories, one is communication failure, and the other is subsystem function failure. Each controlled node detects the communication status between itself and the main control node, judges the communication fault, and the controlled node adopts the processing method of fault passive processing to handle the fault passively. Among them, the judgment method and judgment process adopted are the same as the above-mentioned judgment process. For example, if the propulsion motor driver detects a communication failure, it cannot be controlled by the main control module at this time, and the power subsystem can automatically stop according to needs; If under-voltage, over-current, over-temperature and other out-of-range functional faults are detected, the fault needs to be reported, and the main control node is notified to record the fault source and actively handle the fault.

Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than limit them. Although the present invention has been described in detail with reference to the embodiments, those skilled in the art should understand that modifications or equivalent replacements to the technical solutions of the present invention do not depart from the spirit and scope of the technical solutions of the present invention, and all of them should be included in the scope of the present invention. within the scope of the claims.

Claims (3)

1. An underwater vehicle control system fault detection and fault handling method, the control system comprising: a master control node and a plurality of controlled nodes; the master control node and the controlled node carry out master-slave communication through a digital bus; performing fault detection and fault processing on the internal communication of the control system and each controlled node at the same time;

aiming at a certain controlled node arranged in the underwater vehicle, the main control node is communicated with the controlled node internally through a digital bus, and the method is adopted to detect and process faults of the controlled node;

the method comprises the following steps:

the master control node sends a broadcasting instruction of a period T to the controlled node through a digital bus;

the main control node judges and generates a corresponding fault record according to the response condition and response content of the controlled node, and stores the fault record as record-level fault data to finish fault detection of the controlled node;

the main control node maps the record level fault data to the corresponding processing level fault data in real time according to an externally issued debugging and real-time navigation conversion instruction, and the main control node adopts a fault active processing method to process the fault according to the content of the processing level fault data;

the controlled node detects the communication state of the controlled node and the main control node, judges the communication fault, and carries out the passive fault processing by the controlled node;

the method also comprises the following steps before the implementation of the method: defining a 32-bit CER variable, a 32-bit UER variable, a 32-bit DER variable, a 32-bit CED variable, a 32-bit UED variable, and a 32-bit DED variable;

a CER variable of 32 bits as a buffer for record level communication failures;

a 32-bit UER variable as a buffer for the recording-level data update exception fault;

a 32-bit DER variable serving as a buffer for recording-level data abnormal faults;

the CED variable of 32 bits is used as a cache of the processing-level communication fault, and the processing-level mapping of the CER variable is corresponding;

a 32-bit UED variable is used as a buffer for updating abnormal faults of the processing level data, and the processing level mapping of the corresponding UER variable is performed;

the DED variable of 32 bits is used as a cache of the abnormal faults of the processing level data, and the processing level mapping of the DER variable is corresponding;

the data bits in the 32-bit CER variable are mapped one-to-one with the data bits in the 32-bit CED variable;

the data bits in the 32-bit UER variable are mapped one-to-one with the data bits in the 32-bit UED variable;

the data bits in the 32-bit DER variable are mapped one-to-one with the data bits in the 32-bit DED variable;

resetting each of the different variables to 0 during initialization;

the main control node judges and generates a corresponding fault record according to the response condition and response content of the controlled node, and stores the fault record as record-level fault data to finish fault detection of the controlled node; the specific process is as follows:

the main control node judges and generates a communication fault record according to the response condition of the controlled node, and stores the communication fault record as record-level communication fault data to determine the fault type as a communication fault;

the main control node judges and generates a function fault record according to the response content of the controlled node, and stores the function fault record as record-level function fault data to determine the type of the fault as a function fault;

the main control node judges and generates a communication fault record according to the response condition of the controlled node, and stores the communication fault record as record-level communication fault data to determine the fault type as a communication fault; the specific process is as follows:

the master control node sends a broadcasting instruction with the period of T to the controlled node, and the controlled node responds;

if the response of the controlled node is detected in the period T, the main control node controls the value of the communication anomaly counter corresponding to the controlled node to be clear 0;

if the response of the controlled node is not detected in the period T, the main control node controls the value of the communication anomaly counter corresponding to the controlled node to be increased by 1;

counting the numerical value of the communication anomaly counter, recording the numerical value as a statistical value, and comparing the statistical value with a preset threshold value;

if the statistical value is greater than or equal to a preset threshold value, setting the bit value of the data bit corresponding to the controlled node in the CER variable of 32 bits to be 1, generating a communication fault record, updating the record-level communication fault data, storing, and determining the fault type as a communication fault;

if the statistic value is smaller than a preset threshold value, not performing any processing;

the main control node judges and generates a function fault record according to the response content of the controlled node, and stores the function fault record as record-level function fault data to determine the type of the fault as a function fault; the specific process is as follows:

the master control node sends a broadcasting instruction with the period of T to the controlled node, and the controlled node responds;

if the response of the controlled node is detected in the period T, recording the response content of the controlled node in the current period as the function data of the response, recording the response content as the function data of the current period, and judging whether the function data of the current period is the same as the function data of the previous period;

if the functional data in the current period is different from the functional data in the previous period, the main control node controls the numerical value of the data refreshing exception counter corresponding to the controlled node to be 0;

if the function data of the current period is the same as the function data of the previous period, the main control node controls the value of the data refreshing abnormal counter corresponding to the controlled node to be increased by 1;

counting the numerical value of the data refreshing abnormal counter, recording the numerical value as a refreshing statistical value, and comparing the refreshing statistical value with a preset refreshing threshold value;

if the refresh statistic value is smaller than a preset refresh threshold value, not performing any processing;

if the refresh statistical value is greater than or equal to a preset refresh threshold value, setting the bit value of the data bit corresponding to the controlled node in the 32-bit UER variable to be 1, generating a data refresh abnormal fault record, updating the data refresh fault data of the record data, storing, and determining the fault type as the data refresh abnormal in the functional fault;

meanwhile, judging whether the functional data of the current period exceeds a preset boundary range or not;

if the functional data in the current period exceeds the preset boundary range, setting the bit value of the data bit corresponding to the controlled node in the 32-bit DER variable to be 1, generating a data abnormal fault record, updating the record-level data abnormal fault data, storing, and determining the fault type as the data abnormality in the functional fault;

if the functional data of the current period does not exceed the preset boundary range, not performing any processing;

the method further comprises the steps of: each controlled node detects the communication state between the controlled node and the main control node, judges and generates a communication fault record according to the detected receiving condition of the main control node periodically sending a broadcast instruction, stores the communication fault record on the controlled node, and determines the fault type as a communication fault; the controlled node carries out automatic processing according to the self function type, and the fault processing process does not have the participation of the main control node, namely the fault is passively processed.

2. The method for detecting and processing faults of an underwater vehicle control system according to claim 1, wherein the main control node maps recording level fault data to corresponding processing level fault data in real time according to an externally issued command for debugging and turning to actual navigation, and the main control node adopts a fault active processing method to process faults according to the content of the processing level fault data; the specific process is as follows:

when the external issuing of the instruction to debug the transition to real navigation, each cycle thereafter operates as follows:

updating the record level communication fault data copy to the processing level communication fault data, in particular updating the data bit copy of the CER variable to the data bit of the CED variable;

updating the record level data refresh fault data copy to the processing level data refresh fault data, in particular updating the data bit copy of the UER variable to the data bit of the UED variable;

copying and updating the recording-level data abnormal fault data to the processing-level data abnormal fault data, and particularly copying the data bit update of the DER variable to the data bit of the DED variable;

if the data bit of the CED variable, the data bit of the UED variable or the arbitrary position of the data bit of the DED variable is larger than 0, adopting a fault active processing method to process the fault;

if all positions of the data bit of the CED variable, the data bit of the UED variable, or the data bit of the DED variable are equal to 0, no processing is performed.

3. The method for detecting and processing the faults of the control system of the underwater vehicle according to claim 1, wherein the specific process of judging the communication abnormality of each controlled node is as follows:

a timer with the same period T as the broadcasting instruction of the main control node is arranged in the controlled node;

if the controlled node detects a main control node broadcasting instruction in a timer period T, the value of a communication anomaly counter in the controlled node is clear 0;

if the controlled node does not detect the main control node broadcasting instruction in the timer period T, the value of a communication anomaly counter in the controlled node is increased by 1;

counting the numerical value of the communication anomaly counter, recording the numerical value as a statistical value, and comparing the statistical value with a preset threshold value;

if the statistical value is greater than or equal to the preset threshold value, recording communication faults and performing fault passive processing

If the statistic is smaller than the preset threshold value, no processing is performed.