Functional safety control system and fault diagnosis decision-making system of electric forklift

Abstract

The invention discloses a function safety control system and a fault diagnosis decision-making system of an electric forklift, which belong to the technical field of safety control systems of electric forklifts and comprise safety function modules of a driving control subsystem, an operation position monitoring subsystem, a starting accident acceleration subsystem and a vehicle driving accident behavior subsystem. And the safety function module executes the safety control logic of each subsystem. The system has the advantages that a function safety control system and a fault diagnosis decision-making system applied to the electric forklift are built; the problems of safety function protection, fault tracing after function protection and fault decision processing and repairing of the electric forklift during running control state monitoring, operation position monitoring, flat road starting accident and accident behavior monitoring during vehicle running are solved.

Description

Technical Field

[0001] The present invention relates to the technical field of functional safety control systems for electric forklifts, and in particular to a functional safety control system and a fault diagnosis decision system for electric forklifts. Background Art

[0002] With the increasing global awareness of environmental protection and the requirements of various countries to reduce carbon emissions, electric forklifts have become an ideal substitute for traditional internal combustion forklifts due to their zero emissions and low noise, with low maintenance costs and energy consumption. The application scenarios of electric forklifts have expanded from traditional warehouse handling to more industries, such as ports, automobiles, steel and other industries. As the demand for overseas export business of electric forklifts gradually increases, the overseas market has higher and higher requirements for the functional safety compliance of electric forklifts.

[0003] Technological advances in electronic control systems, fault diagnosis systems, and post-fault processing decision systems have made the demand for electric forklifts not only urgent in terms of functional safety development, but also in terms of fault model building and post-fault processing decision-making. Summary of the invention

[0004] The present invention aims to solve the technical problem mentioned in the above background technology. In a first aspect, a functional safety control system of an electric forklift is provided to facilitate the detection of the safety of various systems of the electric forklift.

[0005] The second aspect provides a fault diagnosis decision system, which is used in combination with the functional safety control system of the electric forklift provided by the first aspect, and uses its detection data to calculate the most likely cause of the electric forklift failure, making it convenient for authorized technicians to perform fault recovery.

[0006] Based on the above objectives, the technical solution provided in the first aspect of the present invention is: a functional safety control system for an electric forklift, including safety function modules for a drive control subsystem, an operating position monitoring subsystem, a starting unexpected acceleration subsystem, and a vehicle driving unexpected behavior subsystem, and the safety function modules execute the safety control logic of each subsystem.

[0007] In some embodiments, the safety control logic of each subsystem is as follows: after the vehicle is powered on and initialized, the safety function module monitors the signal status of each component of the drive control subsystem, the operating position monitoring subsystem, the start-up accidental acceleration subsystem, and the vehicle driving accidental behavior subsystem. The detection cycle of the signal of each component of the safety function module monitoring subsystem is ≤50ms. If the abnormal trigger condition defined by the monitoring is triggered, the safety function protection state of the corresponding subsystem module is entered. At this time, the main positive contactor of the BMS is disconnected, and an authorized technician is required to perform fault recovery;

[0008] If no abnormality is detected or the technician has restored the fault, the signals of each component of the drive control subsystem, the operating position monitoring subsystem, the start-up unexpected acceleration subsystem, and the vehicle driving unexpected behavior subsystem will continue to be monitored.

[0009] In some embodiments, the safety function module monitors the signals of various components of the drive control subsystem, the operating position monitoring subsystem, the start-up accidental acceleration subsystem, and the vehicle driving accidental behavior subsystem, including the combination handle, the accelerator pedal, the brake pedal, the seat switch, the handbrake switch, the key switch, and the emergency stop switch, and determines whether to enter the corresponding subsystem safety function protection state based on the set monitoring signal threshold.

[0010] In some embodiments, when the vehicle is started, the events that the drive control subsystem needs to detect are:

[0011] Trigger event 1: forward gear switch is TRUE;

[0012] Trigger event 2: the reverse gear switch is TRUE;

[0013] Trigger event 3: There is output from the accelerator pedal;

[0014] Fault detection trigger events: sensor failure, controller failure, actuator failure.

[0015] In some embodiments, the triggering events that the operating position monitoring subsystem needs to detect are:

[0016] Trigger event 1: The accelerator pedal is always in output and the drive motor does not decelerate normally 2 seconds after the operator leaves the seat;

[0017] Trigger event 2: There is no output from the accelerator pedal and the drive motor does not decelerate normally 0.2s after the operator leaves the seat;

[0018] Trigger event 3: There is no reset action during deceleration, and the deceleration state of the drive motor changes;

[0019] Trigger event 4: There is no reset action after standstill, and the standstill state of the drive motor changes.

[0020] Trigger event 5: Fault detection trigger event. Fault detection trigger events include sensor fault, controller fault, and actuator fault.

[0021] In some embodiments, the triggering events that need to be detected by the startup unintended acceleration subsystem are:

[0022] Trigger event 1: driving direction and gear position do not match;

[0023] Trigger event 2: There is no speed request, and the acceleration exceeds the limit;

[0024] Trigger event 3: Fault detection trigger event. Fault detection trigger events include sensor fault, controller fault, and actuator fault.

[0025] In some embodiments, the triggering events that the vehicle driving unexpected behavior subsystem needs to detect are:

[0026] Trigger event 1: Driving direction response trigger conditions:

[0027] When in forward gear, the motor speed is negative and does not decelerate;

[0028] When in reverse gear, the motor speed is positive and not decelerating;

[0029] When in neutral, the motor speed is unstable and does not slow down;

[0030] Trigger event 2: unexpected acceleration trigger event, that is, the absolute value of the motor acceleration is greater than 1.3 times the maximum acceleration and is not in a deceleration state;

[0031] Trigger event 3: unexpected speed trigger event, that is, the absolute value of the motor speed is greater than 1.2 times the absolute value of the expected speed and is not in a deceleration state;

[0032] Trigger event 4: Fault detection trigger event. Fault detection trigger events include sensor fault, controller fault, and actuator fault.

[0033] The second aspect of the present invention provides a fault diagnosis decision system, which is used based on the functional safety control system of the electric forklift provided in the first aspect, collects historical fault model data samples of the drive control subsystem, the operating position monitoring subsystem, the start-up accident acceleration subsystem, and the vehicle driving accident behavior subsystem, and uses the Bayesian greedy algorithm to perform model training and establish a fault diagnosis inference model for each subsystem;

[0034] According to the set of fault probabilities corresponding to the potential causes inferred by the fault diagnosis model, the maximum likelihood estimation algorithm is used to arrange the probability of potential fault causes, establish the causal relationship between historical fault data, and represent the model through a directed acyclic graph;

[0035] The potential causes of triggering the safety status protection are recommended to the authorized technicians one by one according to the ranking values ​​of the probability values. The authorized technicians will handle and repair the faults according to the probability ranking causes recommended by the diagnostic model. After the repair, normal operation will be restored and the safety status protection will be released.

[0036] Furthermore, the method of establishing the fault diagnosis reasoning model of each subsystem includes:

[0037] S1. Construct the fault set of each functional safety subsystem, respectively , , , express:

[0038] ; ;

[0039] ; ;

[0040] In the collection, , b…is the triggering event of the safety function of subsystem 1, ...is the safety function of subsystem 2 ...is the triggering event for the safety function of subsystem 3, …trigger events for the safety functions of subsystem 4.

[0041] S2. Collect and summarize the reasons that cause each functional safety subsystem to enter a safe state protection state. express:

[0042]

[0043] , , ;

[0044] In the collection, , …are potential causes of triggering the safety function triggering event of safety subsystem 1; , …are potential causes of triggering the safety function triggering event of safety subsystem 2; …are potential causes of triggering the safety function triggering event of safety subsystem 3; , ...are the potential causes that lead to the triggering of the safety function triggering event of the safety subsystem 4. Among them, the cause set for the safety state protection Expandable at any time.

[0045] S3. Using the Bayesian greedy learning algorithm to train the quantified historical fault cause data sample set collected from each functional safety subsystem, the Bayesian fault diagnosis model can be obtained respectively:

[0046] ;

[0047] Where: Represents the directed acyclic graph structure DAG of the Bayesian fault diagnosis model; Conditional probability table representing the variables;

[0048] The trained Bayesian fault diagnosis model is evaluated using the BIC scoring function, and the optimal Bayesian fault diagnosis model obtained through training is saved. The BIC scoring function is as follows:

[0049] ;

[0050] ;

[0051] Where: is the maximum likelihood value of the model on the given data, that is, the likelihood function value after the model parameters are estimated. is the number of free parameters in the model, that is, the complexity of the model, The dimension of the network structure representing the Bayesian diagnostic model;

[0052] The model with the smallest BIC value is selected as the final Bayesian fault diagnosis model of the functional safety subsystem.

[0053] Furthermore, according to the set of fault probabilities corresponding to the potential causes inferred by the fault diagnosis model, the method of using the maximum likelihood estimation algorithm to arrange the probability of the potential fault causes includes:

[0054] When the subsystem triggers the safety function, the Bayesian fault diagnosis model of the current subsystem is switched , using the maximum likelihood estimation algorithm to Each set of fault causes, enters the potential causes of safe state protection for diagnostic reasoning;

[0055] S1. First, construct the likelihood function based on independent observation data:

[0056] ;

[0057] Where: is the parameter to be estimated, is the observed sample, is the likelihood function;

[0058] S2. Take the logarithm of the likelihood function and obtain the log-likelihood function:

[0059] ;

[0060] S3. According to the objective requirement of the parameter value function with the maximum log-likelihood, the formula can be obtained as follows:

[0061] ;

[0062] right about Take the derivative and set it to 0:

[0063] ;

[0064] S4. By solving the equation, we can get the parameters Estimated value of , and then reverse reasoning to get the set of fault causes The corresponding probability value is express:

[0065] ;

[0066] ;

[0067] ;

[0068] ;

[0069] S5. After obtaining the probability of potential fault causes using the fault diagnosis model, enter the fault post-processing decision module and sort the probability values ​​of various fault causes for the touch operation position monitoring to enter the safe state protection from large to small;

[0070] The potential causes of triggering the safety status protection are recommended to the authorized technicians one by one according to the ranking values ​​of the probability values. The authorized technicians handle and repair the faults according to the probability ranking fault causes recommended by the diagnostic model.

[0071] The advantages of the present invention compared with the prior art are: a functional safety control system and a fault diagnosis decision system applied to electric forklifts are established, which solves the problems of safety function protection of electric forklifts in driving control status monitoring, operating position monitoring, accidents when starting on flat roads, and accidental behavior monitoring during vehicle driving, fault tracing after functional protection, and fault decision processing and repair.

[0072] Additional aspects and advantages of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] Figure 1 It is a schematic diagram of the overall logic flow of the electric forklift functional safety control system and the fault diagnosis decision system according to an embodiment of the present invention;

[0074] Figure 2 It is a schematic diagram of the overall process framework of the fault diagnosis decision of the safety control modules of each subsystem of the electric forklift according to the embodiment of the present invention;

[0075] Figure 3 yes Figure 2 Schematic diagram of the fault diagnosis decision process of the drive control subsystem shown in;

[0076] Figure 4yes Figure 2 A schematic diagram of the fault diagnosis decision process of the operating position monitoring subsystem shown in FIG.

[0077] Figure 5 yes Figure 2 The schematic diagram of the decision-making process of fault diagnosis of the unintentional acceleration subsystem when starting from a flat road is shown in FIG.

[0078] Figure 6 yes Figure 2 Schematic diagram of the fault diagnosis decision process for the vehicle unexpected behavior subsystem shown in FIG. DETAILED DESCRIPTION

[0079] The present invention is described in further detail below.

[0080] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. The following embodiments are used to illustrate the present invention but are not used to limit the scope of the present invention.

[0081] Combination Figure 1 , Figure 2 As shown, this embodiment illustrates the combined use of the electric forklift functional safety control system and the fault diagnosis decision system. The electric forklift functional safety control system performs Bayesian fault diagnosis after triggering the safety function, and makes a fault handling decision after fault tracing. The electric forklift functional safety control system is further subdivided into a drive control subsystem, an operating position monitoring subsystem, a flat road starting accident acceleration subsystem, and a vehicle driving accident behavior subsystem.

[0082] After the vehicle system is powered on and initialized, key input signals such as the combination handle, accelerator pedal, brake pedal, seat switch, handbrake switch, key switch, and emergency stop switch are collected in real time. At the same time, the safety function modules of each subsystem are started to monitor the safety status of the vehicle. If it is detected that one of the signals is abnormal and causes the safety function to be triggered, the safety status protection is performed. After entering the status protection, the constructed subsystem Bayesian fault diagnosis model is entered respectively to perform reverse diagnostic reasoning on the possible fault causes that trigger the safety function. The Bayesian Maximum Likelihood Estimation (MLE) is used to trace a group of potential fault causes that trigger the safety function. The fault processing decision module is used to perform decision analysis on a set of fault causes after reverse tracing reasoning, and finally the most likely fault is obtained. This most likely fault is recommended to the authorized technician to facilitate the fault processing and repair and safety function reset. At the same time, it is monitored whether the authorized technician has performed fault recovery. If the authorized technician completes the fault recovery, the safety protection state is released and the signal monitoring module is entered.

[0083] Combination Figure 3 As shown in the figure, the safety function logic and implementation process of the drive control subsystem are as follows:

[0084] (1) Safety function logic design:

[0085] The safety controller operation cycle is set to ≤50ms. If the safety function triggering condition is detected at startup, the controller will enter the safe state and the main positive contactor of the BMS will be disconnected.

[0086] (2) Safety function triggering conditions:

[0087] The events that need to be detected when the vehicle starts are as follows:

[0088] a: Trigger event 1: The forward gear switch is TRUE;

[0089] b: Trigger event 2: The reverse gear switch is TRUE;

[0090] c: Trigger event 3: the accelerator pedal has output;

[0091] d: Fault detection trigger events: sensor failure, controller failure, actuator failure.

[0092] If one of the events is detected when the vehicle is started, it is judged that the safety function of the drive control system is triggered. In order to prevent the forklift from suddenly moving and causing damage to the operator and surrounding personnel, resulting in danger, it should enter the safe state:

[0093] (3) Design of fault diagnosis model for driving control subsystem:

[0094] If a safety function setting event is triggered and the safety state is entered, while the safety control function is enabled for action protection, it is necessary to further diagnose the cause of the fault. This embodiment uses Bayesian training on the historical fault data of each working device (i.e., each component) to obtain a fault diagnosis model, and reversely infers the fault to obtain the cause of the fault in order to resolve the fault.

[0095] S1. Construct the driving control subsystem fault set, and use express:

[0096] ;

[0097] In the collection, Trigger event for safety function 1, Trigger event 2 for safety function, Safety function triggering event 3, Trigger event for fault detection.

[0098] S2. Collect and summarize the reasons that cause the driving control subsystem to enter the safety state protection state. express:

[0099] ;

[0100] In the collection, Potential causes of triggering event 1: forward gear switch is TRUE. Potential reasons for triggering event 2, the reverse gear switch is TRUE, Potential causes of triggering event 3: accelerator pedal output. Potential causes of related sensor failures, controller failures, and actuator failures. A collection of reasons for safe state protection Expandable at any time.

[0101] S3. Use the Bayesian greedy learning algorithm to train the collected and quantified historical fault cause data sample set to obtain the Bayesian fault diagnosis model:

[0102] ;

[0103] Where: Represents the directed acyclic graph structure DAG (Directed Acyclic Graph) of the Bayesian fault diagnosis model; Represents the conditional probability table of the variables.

[0104] The trained Bayesian fault diagnosis model is evaluated using the BIC (Bayesian Information Cri-teria) scoring function, and the optimal Bayesian fault diagnosis model obtained through training is saved. The BIC scoring function is shown below:

[0105] ;

[0106] ;

[0107] Where: is the maximum likelihood value of the model on the given data, that is, the likelihood function value after the model parameters are estimated. is the number of free parameters in the model, that is, the complexity of the model, Represents the network structure dimension of the Bayesian diagnostic model.

[0108] because It measures how well the model fits the data. It is a penalty term for model complexity. The larger this part is, the more complex the model is. Therefore, the model with the smallest BIC value is selected as the final Bayesian fault diagnosis model for the driving control subsystem.

[0109] (4) Post-fault processing decision and repair module design

[0110] When the safety function is triggered, the Bayesian fault diagnosis model of the driving control subsystem is switched , using the maximum likelihood estimation algorithm to Diagnose potential reasons for entering the safe state protection;

[0111] S1. First, construct the likelihood function based on independent observation data:

[0112] ;

[0113] Where: is the parameter to be estimated, is the observed sample, is the likelihood function.

[0114] S2. To simplify the calculation, take the logarithm of the likelihood function and obtain the log-likelihood function:

[0115] ;

[0116] S3. According to the objective requirement of the parameter value function with the maximum log-likelihood, the formula can be obtained as follows:

[0117] ;

[0118] right about Take the derivative and set it to 0:

[0119]

[0120] S4. By solving the equation, we can get the parameters Estimated value of , and then reverse reasoning to get the set of fault causes The corresponding probability value is express:

[0121] ;

[0122] S5. After using the diagnostic model to obtain the probability of potential fault causes, enter the fault post-processing decision module and calculate the probability value of each fault cause. The potential causes of triggering the safety status protection are sorted from large to small, and are recommended to authorized technicians one by one according to the sorted probability values. The authorized technicians handle and repair the faults according to the probability sorted fault causes recommended by the diagnostic model.

[0123] (5) Safety status reset conditions

[0124] Failure recovery:

[0125] When a fault is detected that triggers the safety function and the controller loop enters a safe state, the fault should be repaired by an authorized technician. The authorized technician will sort the fault causes according to the probability recommended by the diagnostic model and perform fault processing and repair to restore normal operation.

[0126] Combination Figure 4 As shown, the operating position monitoring safety function

[0127] (1) Safety function logic:

[0128] The components related to the operating position monitoring safety function include: seat switch, combination handle, accelerator pedal, and the safety controller operation cycle is set to ≤50ms. If the safety function triggering conditions are detected at startup, the controller will enter the safe state and the main positive contactor of the BMS will be disconnected.

[0129] (2) Safety function triggering events:

[0130] a: Trigger event 1: The accelerator pedal is always in output and the drive motor does not decelerate normally 2 seconds after the operator leaves the seat;

[0131] b: Trigger event 2: There is no output on the accelerator pedal and the drive motor does not decelerate normally 0.2s after the operator leaves the seat;

[0132] c: Trigger event 3: There is no reset action during deceleration, and the deceleration state of the drive motor changes;

[0133] d: Trigger event 4: There is no reset action after standstill, and the standstill state of the drive motor changes.

[0134] e: Trigger event 5 Fault detection trigger event:

[0135] Trigger event 1 sensor failure;

[0136] Trigger event 2 controller failure;

[0137] Triggering event 3 actuator failure.

[0138] If one of them is triggered, it means that the safety function of the drive control system is triggered. In order to prevent the forklift from suddenly moving and causing damage to the operator and surrounding personnel, resulting in danger, the safe state should be entered:

[0139] (3) Design of operating position monitoring fault diagnosis model:

[0140] If the operation position monitoring safety function setting event is triggered and the safe state is entered, while the safety control function is enabled for action protection, it is necessary to further diagnose the cause of the fault. This embodiment uses Bayesian training on the historical fault data of the operation position state protection to obtain a fault diagnosis model, and reversely infers the fault to obtain the cause of the fault in order to resolve the fault.

[0141] S1, build the operation location subsystem fault set, use express:

[0142] ;

[0143] In the collection, Trigger event for safety function 1, Trigger event 2 for safety function, Safety function triggering event 3, Safety function triggering event 4, Trigger event for fault detection.

[0144] S2. Collect and summarize the reasons that cause the operating position subsystem failure to enter the safe state protection, and use the collection express:

[0145] ;

[0146] In the collection, The potential cause of triggering event 1 is that the accelerator pedal is always on and the drive motor does not decelerate normally 2 seconds after the operator leaves the seat. Potential reasons for triggering event 2: the accelerator pedal has no output and the drive motor does not decelerate normally 0.2s after the operator leaves the seat. Potential reasons for triggering event 3 (no reset action during deceleration) and driving motor deceleration state change: To cause no reset action after the trigger event 4 stops, the drive motor stops state changes, Potential causes of triggering events for fault detection. Set of causes for safe state protection Expandable at any time.

[0147] S3. Use the Bayesian greedy learning algorithm to train the collected and quantified historical fault cause data sample set to obtain the Bayesian fault diagnosis model:

[0148] ;

[0149] Where: Represents the directed acyclic graph structure DAG (Directed Acyclic Graph) of the Bayesian fault diagnosis model; Represents the conditional probability table of the variables.

[0150] The trained Bayesian fault diagnosis model is evaluated using the BIC (Bayesian Information Cri-teria) scoring function, and the optimal Bayesian fault diagnosis model obtained through training is saved. The BIC scoring function is shown below:

[0151]

[0152]

[0153] Where: is the maximum likelihood value of the model on the given data, that is, the likelihood function value after the model parameters are estimated. is the number of free parameters in the model, that is, the complexity of the model, Represents the network structure dimension of the Bayesian diagnostic model.

[0154] because It measures how well the model fits the data. It is a penalty term for model complexity. The larger this part is, the more complex the model is. Therefore, the model with the smallest BIC value is selected as the final Bayesian fault diagnosis model for the operating position monitoring subsystem.

[0155] (4) Post-fault processing decision and repair module design

[0156] When the safety function is triggered, the Bayesian fault diagnosis model of the operating position monitoring subsystem is switched on , using the maximum likelihood estimation algorithm to Diagnose potential reasons for entering the safe state protection;

[0157] S1. First, construct the likelihood function based on independent observation data:

[0158]

[0159] Where: is the parameter to be estimated, is the observed sample, is the likelihood function.

[0160] S2. To simplify the calculation, take the logarithm of the likelihood function and obtain the log-likelihood function:

[0161]

[0162] S3. According to the objective requirement of the parameter value function with the maximum log-likelihood, the formula can be obtained as follows:

[0163]

[0164] right about Take the derivative and set it to 0:

[0165]

[0166] S4. By solving the equation, we can get the parameters Estimated value of , and then reverse reasoning to get the set of fault causes The corresponding probability value is express:

[0167]

[0168] S5. After obtaining the probability of potential fault causes using the fault diagnosis model, enter the fault post-processing decision module to monitor the operation position of the touch screen and enter the safe state protection. The probability value of each fault cause , , , , The potential causes of triggering the safety status protection are sorted from large to small, and are recommended to authorized technicians one by one according to the sorted probability values. The authorized technicians handle and repair the faults according to the probability sorted fault causes recommended by the diagnostic model.

[0169] (5) Safety status reset conditions

[0170] Failure recovery:

[0171] When the operation position monitoring safety function is triggered to enter the state protection and the controller loop enters the safe state, the fault should be repaired by an authorized technician. The authorized technician will sort the fault causes according to the probability recommended by the diagnostic model, and then handle and repair the fault to restore normal operation.

[0172] Combination Figure 5 As shown in the figure, the safety function logic and implementation process of the unintentional acceleration subsystem when starting on a flat road;

[0173] (1) Safety function logic of the accidental acceleration subsystem when starting on a flat road:

[0174] The safety function of the unexpected acceleration subsystem when starting on a flat road is composed of a combination handle and coded safety-related components. The controller operation cycle is set to ≤50ms. After starting, if the driving direction is detected to be inconsistent with the gear direction, the safety function will be triggered; if the acceleration exceeds the limit when there is no speed request, the safety function will also be triggered. After the safety function is triggered, the controller will enter a safe state, and the main positive contactor of the BMS will be disconnected for safety protection.

[0175] (2) Safety function triggering events:

[0176] a: Trigger event 1 Driving direction and gear position do not match:

[0177] That is, the safety function of unexpected acceleration at start-up - uncontrolled starting direction is triggered to prevent the forklift from losing control while moving and causing harm to personnel and causing danger.

[0178] b: Trigger event 2: No speed request, acceleration exceeds the limit:

[0179] That is, the safety function of unexpected acceleration at start-up - uncontrolled acceleration at start-up is triggered to prevent the forklift from losing control while moving and causing harm to personnel and causing danger.

[0180] c: Trigger event 3 Fault detection trigger event:

[0181] Sensor failure;

[0182] Controller failure;

[0183] Actuator failure;

[0184] If one of them is triggered, it means that the safety function of the drive control system is triggered. In order to prevent the forklift from suddenly moving and causing damage to the operator and surrounding personnel, resulting in danger, the safe state should be entered:

[0185] (3) Design of fault diagnosis model for the subsystem of unexpected acceleration during flat road start:

[0186] If the safety function setting event of the unexpected acceleration subsystem from level start is triggered and the system enters the safe state, while enabling the safety control function for action protection, it is necessary to further diagnose the cause of the fault. This embodiment uses Bayesian training on the historical fault data of the unexpected acceleration subsystem from level start to obtain a fault diagnosis model, and reversely infers the fault to obtain the cause of the fault in order to resolve the fault.

[0187] S1. Construct the fault set of the subsystem of unexpected acceleration when starting on a flat road. express:

[0188]

[0189] In the collection, Trigger event for safety function 1, Trigger event 2 for safety function, The safety function triggers event 3.

[0190] S2. Collect and summarize the reasons for the failure of the subsystem to enter the safety state protection when starting from flat road. express:

[0191]

[0192] In the collection, Potential causes of triggering event 1: driving direction and gear mismatch, Potential causes of triggering event 2: no speed request, acceleration exceeding the limit, Detect potential causes of trigger events for the fault that leads to trigger event 3. Set of causes for safe state protection Expandable at any time.

[0193] S3. Use the Bayesian greedy learning algorithm to train the collected and quantified historical fault cause data sample set to obtain the Bayesian fault diagnosis model:

[0194]

[0195] Where: Represents the directed acyclic graph structure DAG (Directed Acyclic Graph) of the Bayesian fault diagnosis model; Represents the conditional probability table of the variables.

[0196] The trained Bayesian fault diagnosis model is evaluated using the BIC (Bayesian Information Cri-teria) scoring function, and the optimal Bayesian fault diagnosis model obtained through training is saved. The BIC scoring function is shown below:

[0197]

[0198]

[0199] Where: is the maximum likelihood value of the model on the given data, that is, the likelihood function value after the model parameters are estimated. is the number of free parameters in the model, that is, the complexity of the model, Represents the network structure dimension of the Bayesian diagnostic model.

[0200] because It measures how well the model fits the data. It is a penalty term for model complexity. The larger this part is, the more complex the model is. Therefore, the model with the smallest BIC value is selected as the final Bayesian fault diagnosis model for the operating position monitoring subsystem.

[0201] (4) Post-fault processing decision and repair module design

[0202] When the safety function is triggered, the Bayesian fault diagnosis model of the accidental acceleration subsystem of flat road start is cut in , using the maximum likelihood estimation algorithm to Diagnose potential reasons for entering a safe state of protection

[0203] First, construct the likelihood function based on independent observation data:

[0204]

[0205] Where: is the parameter to be estimated, is the observed sample, is the likelihood function.

[0206] To simplify the calculation, take the logarithm of the likelihood function and obtain the log-likelihood function:

[0207]

[0208] According to the objective requirement of the parameter value function with the maximum log-likelihood, the formula can be obtained as follows:

[0209]

[0210] right about Take the derivative and set it to 0:

[0211]

[0212] By solving the equation, we can get the parameters Estimated value of , and then reverse reasoning to get the set of fault causes As the corresponding probability value, use express:

[0213]

[0214] After using the diagnostic model to obtain the probability of potential fault causes, enter the fault post-processing decision module and calculate the probability value of each fault cause. The potential causes of triggering the safety status protection are sorted from large to small, and are recommended to authorized technicians one by one according to the sorted probability values. The authorized technicians handle and repair the faults according to the probability sorted fault causes recommended by the diagnostic model.

[0215] (5) Safety status reset conditions

[0216] Failure recovery:

[0217] When the safety function of the unexpected acceleration subsystem when starting on a flat road enters the state protection and the controller loop enters the safe state, the fault should be repaired by an authorized technician. The authorized technician will handle and repair the fault according to the probability ranking of the fault causes recommended by the diagnostic model to restore normal operation.

[0218] Combination Figure 6 As shown, the vehicle driving unexpected behavior subsystem safety function

[0219] (1) Logic of the safety function for unexpected behavior of the vehicle during driving:

[0220] The unexpected behavior safety function of the vehicle while driving is composed of a combination handle and coded safety-related components. The controller operation cycle is set to ≤50ms. If unexpected behaviors such as driving direction reaction, unexpected acceleration, unexpected speed, etc. are detected, the safety function is triggered and the controller will enter a safe state. The main positive contactor of the BMS is disconnected for safety protection.

[0221] (2) Safety function triggering events:

[0222] a: Trigger event 1 Driving direction response trigger condition:

[0223] When in forward gear, the motor speed is negative and does not decelerate;

[0224] When in reverse gear, the motor speed is positive and not decelerating;

[0225] When in neutral, the motor speed is unstable and does not slow down;

[0226] b: Trigger event 2 Unexpected acceleration trigger event:

[0227] The absolute value of the motor acceleration is greater than 1.3 times the maximum acceleration and is not in a deceleration state.

[0228] c: Trigger event 3 Unexpected speed trigger event:

[0229] The absolute value of the motor speed is greater than 1.2 times the absolute value of the expected speed and is not in a deceleration state.

[0230] d: Trigger event 4 Fault detection trigger event:

[0231] Sensor failure;

[0232] Controller failure;

[0233] Actuator failure.

[0234] (3) Design of fault diagnosis model for unexpected behavior of vehicle during driving:

[0235] If the safety function setting event of unexpected behavior while the vehicle is driving is triggered and the vehicle enters a safe state, while enabling the safety control function for action protection, it is necessary to further diagnose the cause of the fault. This embodiment uses Bayesian training on the historical fault data of unexpected behavior while the vehicle is driving to obtain a fault diagnosis model, and reversely infers the fault to obtain the cause of the fault in order to resolve the fault.

[0236] S1, vehicle driving unexpected behavior subsystem fault set, express:

[0237]

[0238] In the collection, Trigger event for safety function 1, Trigger event 2 for safety function, Safety function triggering event 3, Event 4 is triggered for the safety function.

[0239] S2. Collect and summarize the reasons why the vehicle's unexpected behavior subsystem failure enters the safe state protection during driving, and use the collection express:

[0240]

[0241] In the collection, Potential causes of triggering event 1 driving direction response triggering conditions, Potential causes of unexpected acceleration trigger events, Potential causes of the unexpected speed trigger event 3, Trigger event for fault detection. Set of reasons for safe state protection Expandable at any time.

[0242] S3. Use the Bayesian greedy learning algorithm to train the collected and quantified historical fault cause data sample set to obtain the Bayesian fault diagnosis model:

[0243]

[0244] Where: Represents the directed acyclic graph structure DAG (Directed Acyclic Graph) of the Bayesian fault diagnosis model; Represents the conditional probability table of the variables.

[0245] The trained Bayesian fault diagnosis model is evaluated using the BIC (Bayesian Information Cri-teria) scoring function, and the optimal Bayesian fault diagnosis model obtained through training is saved. The BIC scoring function is shown below:

[0246]

[0247]

[0248] Where: is the maximum likelihood value of the model on the given data, that is, the likelihood function value after the model parameters are estimated. is the number of free parameters in the model, that is, the complexity of the model, Represents the network structure dimension of the Bayesian diagnostic model.

[0249] because It measures how well the model fits the data. It is a penalty term for model complexity. The larger this part is, the more complex the model is. Therefore, the model with the smallest BIC value is selected as the final Bayesian fault diagnosis model for the operating position monitoring subsystem.

[0250] (4) Post-fault processing decision and repair module design

[0251] When the safety function is triggered, the Bayesian fault diagnosis model of the vehicle driving unexpected behavior subsystem is switched to , using the maximum likelihood estimation algorithm to Diagnose potential reasons for entering a safe state of protection

[0252] First, construct the likelihood function based on independent observation data:

[0253]

[0254] Where: is the parameter to be estimated, is the observed sample, is the likelihood function.

[0255] To simplify the calculation, take the logarithm of the likelihood function and obtain the log-likelihood function:

[0256]

[0257] According to the objective requirement of the parameter value function with the maximum log-likelihood, the formula can be obtained as follows:

[0258]

[0259] right about Take the derivative and set it to 0:

[0260]

[0261] By solving the equation, we can get the parameters Estimated value of , and then reverse reasoning to get the set of fault causes As the corresponding probability value, use express:

[0262]

[0263] After using the diagnostic model to obtain the probability of potential fault causes, enter the fault post-processing decision module and calculate the probability value of each fault cause. The potential causes of triggering the safety status protection are sorted from large to small, and are recommended to authorized technicians one by one according to the sorted probability values. The authorized technicians handle and repair the faults according to the probability sorted fault causes recommended by the diagnostic model.

[0264] (5) Safety status reset conditions

[0265] Failure recovery:

[0266] When the vehicle's unexpected behavior safety function enters state protection while driving and the controller loop enters a safe state, the fault should be repaired by an authorized technician who will sort the fault causes according to the probability recommended by the diagnostic model and perform fault processing and repair to restore normal operation.

[0267] Although the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the claims and their equivalents.

Claims

1. A functional safety control system for an electric forklift, characterized in that: It includes safety function modules of a drive control subsystem, an operating position monitoring subsystem, a start-up accident acceleration subsystem, and a vehicle driving accident behavior subsystem, and the safety function modules execute the safety control logic of each subsystem.

2. The functional safety control system of the electric forklift according to claim 1 is characterized in that The safety control logic of each subsystem is as follows: after the vehicle is powered on and initialized, the safety function module monitors the signal status of each component of the drive control subsystem, the operating position monitoring subsystem, the start-up accidental acceleration subsystem, and the vehicle driving accidental behavior subsystem. The detection cycle of the signal of each component of the safety function module monitoring subsystem is ≤50ms. If the abnormal trigger condition defined by the monitoring is triggered, the corresponding subsystem module enters the safety function protection state. At this time, the main positive contactor of the BMS is disconnected, and an authorized technician is required to perform fault recovery; If no abnormality is detected or the technician has restored the fault, the signals of each component of the drive control subsystem, the operating position monitoring subsystem, the start-up unexpected acceleration subsystem, and the vehicle driving unexpected behavior subsystem will continue to be monitored.

3. The functional safety control system of the electric forklift according to claim 2, characterized in that: The safety function module monitors the signals of various components of the drive control subsystem, operating position monitoring subsystem, start-up accidental acceleration subsystem, and vehicle driving accidental behavior subsystem, including the combination handle, accelerator pedal, brake pedal, seat switch, handbrake switch, key switch, and emergency stop switch, and determines whether to enter the corresponding subsystem safety function protection state based on the set monitoring signal threshold.

4. The functional safety control system of an electric forklift according to claim 2, characterized in that: When the vehicle is started, the events that the drive control subsystem needs to detect are: Trigger event 1: forward gear switch is TRUE; Trigger event 2: the reverse gear switch is TRUE; Trigger event 3: There is output from the accelerator pedal; Fault detection trigger events: sensor failure, controller failure, actuator failure.

5. The functional safety control system of an electric forklift according to claim 1, characterized in that: The triggering events that the operating position monitoring subsystem needs to detect are: Trigger event 1: The accelerator pedal is always in output and the drive motor does not decelerate normally 2 seconds after the operator leaves the seat; Trigger event 2: There is no output from the accelerator pedal and the drive motor does not decelerate normally 0.2s after the operator leaves the seat; Trigger event 3: There is no reset action during deceleration, and the deceleration state of the drive motor changes; Trigger event 4: No reset action after standstill, the drive motor's standstill state changes; Trigger event 5: Fault detection trigger event. Fault detection trigger events include sensor fault, controller fault, and actuator fault.

6. The functional safety control system of an electric forklift according to claim 1, characterized in that: The triggering events that the startup unexpected acceleration subsystem needs to detect are: Trigger event 1: driving direction and gear position do not match; Trigger event 2: There is no speed request, and the acceleration exceeds the limit; Trigger event 3: Fault detection trigger event. Fault detection trigger events include sensor fault, controller fault, and actuator fault.

7. The functional safety control system of an electric forklift according to claim 1, characterized in that: The triggering events that the vehicle driving unexpected behavior subsystem needs to detect are: Trigger event 1: Driving direction response trigger conditions: When in forward gear, the motor speed is negative and does not decelerate; When in reverse gear, the motor speed is positive and not decelerating; When in neutral, the motor speed is unstable and does not slow down; Trigger event 2: unexpected acceleration trigger event, that is, the absolute value of the motor acceleration is greater than 1.3 times the maximum acceleration and is not in a deceleration state; Trigger event 3: unexpected speed trigger event, that is, the absolute value of the motor speed is greater than 1.2 times the absolute value of the expected speed and is not in a deceleration state; Trigger event 4: Fault detection trigger event. Fault detection trigger events include sensor fault, controller fault, and actuator fault.

8. A fault diagnosis decision system, used in combination with the functional safety control system of an electric forklift according to any one of claims 1 to 7, characterized in that: Collect historical fault model data samples of the drive control subsystem, operating position monitoring subsystem, start-up accidental acceleration subsystem, and vehicle driving accidental behavior subsystem, use the Bayesian greedy algorithm to train the model and establish the fault diagnosis inference model of each subsystem; According to the set of fault probabilities corresponding to the potential causes inferred by the fault diagnosis model, the maximum likelihood estimation algorithm is used to arrange the probability of potential fault causes, establish the causal relationship between historical fault data, and represent the model through a directed acyclic graph; The potential causes of triggering the safety status protection are recommended to the authorized technicians one by one according to the ranking values ​​of the probability values. The authorized technicians will handle and repair the faults according to the probability ranking causes recommended by the diagnostic model. After the repair, normal operation will be restored and the safety status protection will be released.

9. The fault diagnosis decision system according to claim 8, characterized in that: The method for establishing the fault diagnosis abduction model of each subsystem includes: S1. Construct the fault set of each functional safety subsystem, respectively , , , express; ； ； ； ； In the collection, Trigger events for the safety functions of the drive control subsystem, Triggering events for safety functions of the operating position monitoring subsystem; To trigger an event for the safety function of the unexpected acceleration subsystem, Triggering events for safety functions of vehicle unexpected behavior subsystems; S2. Collect and summarize the reasons that cause each functional safety subsystem to enter a safe state protection state. express: 、 、 、 ； In the collection, Potential causes that may lead to the triggering of the event of the drive control subsystem safety function; Potential causes that may lead to triggering of the safety function triggering event of the operating position monitoring subsystem; …are potential causes of the triggering event of the safety function of the Qibu Unintended Acceleration Subsystem; The potential reasons for triggering the event of the vehicle's unexpected driving behavior subsystem safety function triggering; the reasons for the safety state protection are the set Can be expanded at any time; S3. Using the Bayesian greedy learning algorithm to train the quantified historical fault cause data sample set collected from each functional safety subsystem, the Bayesian fault diagnosis model can be obtained respectively: ； Where: Represents the directed acyclic graph structure DAG of the Bayesian fault diagnosis model; Conditional probability table representing the variables; The trained Bayesian fault diagnosis model is evaluated using the BIC scoring function, and the optimal Bayesian fault diagnosis model obtained through training is saved. The BIC scoring function is shown as follows: ； ； Where: is the maximum likelihood value of the model on the given data, that is, the likelihood function value after the model parameters are estimated. is the number of free parameters in the model, that is, the complexity of the model, The dimension of the network structure representing the Bayesian diagnostic model; The model with the smallest BIC value is selected as the final Bayesian fault diagnosis model of the functional safety subsystem.

10. The fault diagnosis decision system according to claim 9, characterized in that: According to the set of fault probabilities corresponding to the potential causes inferred by the fault diagnosis model, the method of using the maximum likelihood estimation algorithm to arrange the probability of potential fault causes includes: When the subsystem triggers the safety function, the Bayesian fault diagnosis model of the current subsystem is switched , using the maximum likelihood estimation algorithm to Each set of fault causes, enters the potential causes of safe state protection for diagnostic reasoning; S1. First, construct the likelihood function based on independent observation data: ； Where: is the parameter to be estimated, is the observed sample, is the likelihood function; S2. Take the logarithm of the likelihood function and obtain the log-likelihood function: ； S3. According to the objective requirement of the parameter value function with the maximum log-likelihood, the formula can be obtained as follows: ； right about Take the derivative and set it to 0: ； S4. By solving the equation, we can get the parameters Estimated value of , and then reverse reasoning to get the set of fault causes The corresponding probability value is express: ； ； ； ； S5. After obtaining the probability of potential fault causes using the fault diagnosis model, enter the fault post-processing decision module and sort the probability values ​​of various fault causes for the touch operation position monitoring to enter the safe state protection from large to small; The potential causes of triggering the safety status protection are recommended to the authorized technicians one by one according to the ranking values ​​of the probability values. The authorized technicians handle and repair the faults according to the probability ranking fault causes recommended by the diagnostic model.

Images