Self-healing fault-tolerant emergency control method for automated container port

Abstract

The application discloses a self-recovery fault-tolerant emergency regulation method for an automated container port. The method comprises the following steps: combing port production operation business association and process paths, and constructing a multi-priority task sequence of a current shift of the port; collecting historical data of a port operation system and equipment, and realizing state feature extraction and hierarchical classification of a port emergency; constructing a port state transition event flow model and a port system T-S fuzzy fault tree, and evaluating and identifying a current port fault state; based on multi-priority task driving of the shift, combining a feedforward-feedback control thought, and according to the identified state of the fault point and emergency information, self-recovery fault-tolerant emergency regulation is carried out at the system and equipment levels. The application effectively improves the self-recovery fault-tolerant capability and online regulation capability of the automated container port, and comprehensively improves the decision-making level, operation efficiency, production income and core competitiveness of the port.

Description

Technical Field

[0001] This invention belongs to the field of emergency control technology for container ports, and particularly relates to a self-healing and fault-tolerant emergency control method for automated container ports. Background Technology

[0002] Due to the high degree of automation in ports, port systems or equipment are unable to respond promptly to emergencies, easily impacting the efficiency and stability of subsequent operations and, in severe cases, disrupting the entire port's production. Specifically, emergencies in automated container ports mainly include equipment damage and performance degradation. The former is primarily affected by factors such as heavy cargo loads and frequent start-stop / reversal operations, leading to major equipment failures and triggering temporary adjustments and overloading issues. The latter is mainly affected by natural environmental factors, such as heavy fog, heavy rain, and strong sunlight, causing sensor malfunctions that significantly affect remote control and positioning, resulting in short-term equipment performance degradation and ultimately leading to overall port performance decline. The probability of these emergencies increases with port operation time, significantly reducing port production efficiency and profitability. Therefore, to improve the self-adjustment and recovery capabilities of automated container ports in response to emergencies, it is urgent to design an efficient self-healing, fault-tolerant emergency control method. Summary of the Invention

[0003] To address the problems existing in the aforementioned background technology, this invention aims to improve the self-adjustment and recovery capabilities of automated container ports in the face of emergencies such as machine damage and equipment performance degradation, minimize the impact of emergencies on the entire port production and operation system, and help automated container ports move towards becoming truly "smart ports." It comprehensively utilizes a series of theoretical methods, including graph theory, event flow, feature analysis, mathematical modeling, fuzzy fault trees, intelligent decision-making, neural networks, and probability statistics, to design a self-healing, fault-tolerant emergency control method for automated container ports, meeting the current operational needs of automated container ports.

[0004] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: a self-healing, fault-tolerant emergency control method for automated container ports, comprising the following steps:

[0005] S1. Analyze the business relationships and technological paths of port production and operation, and construct a multi-priority task sequence for the current work shift at the port;

[0006] S2. Collect historical data from port operation systems and equipment to extract and classify the status characteristics of port emergencies.

[0007] S3. Construct a port state transition event flow model and a port system TS fuzzy fault tree to assess and identify the current port fault state;

[0008] S4. Based on the multi-priority task-driven approach of the work team, combined with the feedforward-feedback control concept, and based on the identified status of the fault point and information on sudden events, self-healing fault-tolerant emergency control is carried out at the system and equipment levels.

[0009] Furthermore, in step S1, the port production and operation business is carried out in units of work shift work plans, which mainly include ship operation plans, barge operation plans, equipment operation plans, loading and unloading site plans, etc.; the process path refers to the process flow and execution sequence and path of related mechanical equipment adopted in the above-mentioned business operations.

[0010] Furthermore, the multi-priority task sequence of the current port shift is constructed based on a real-time task directed graph (DRT), as shown in the following sub-steps:

[0011] S11. First, record the current work shift as... i , sorting out the work team i The relevant tasks and the process path for task execution;

[0012] S12. Secondly, construct the work team. i The set of directed graphs (DRTs) for real-time tasks, denoted as , where the directed graph for any task Characterized as , The set of job nodes representing a directed graph. The set representing the edges between job nodes;

[0013] S13. Subsequently, for the directed graph of the real-time task... Any job node Utilizing ordered pairs It represents the longest execution time (WCET) and the relative time limit (RD) of its operation, while WCET and RD are determined based on port KPIs and historical data, and are constant values;

[0014] S14. Finally, for directed graphs of real-time tasks any edge in The minimum release interval parameter and relative time limit Compare and satisfy <= This is used to control the sequence of tasks, thereby forming work shifts. i Real-time multi-priority task sequences.

[0015] Furthermore, in step S2, the port emergencies include machine damage and equipment performance degradation. Machine damage is mainly affected by adverse working conditions and operational factors, while equipment performance degradation is affected by the natural environment. The historical data of the operating system and equipment is collected through the Terminal Operating System (TOS), Equipment Control System (ECS), and Crane Management System (CMS). The extraction and classification of emergency status features are based on the historical data collected by the system. Starting from the dimensions of emergency type, duration, and impact, principal component analysis (PCA) and K-means clustering algorithm are used to perform feature analysis and extraction in each dimension to quickly achieve the classification and grading of emergencies. Among them, emergency information is represented by multi-dimensional feature sequences.

[0016] Furthermore, in step S3, the assessment and identification of the port fault status specifically includes the following sub-steps:

[0017] S31. Combining the historical data of port systems and equipment and the classification and grading of emergencies collected in step S2, and using data cleaning, data filtering and preprocessing, and causal relationship analysis theories and methods, sort out and construct a port operation system event flow model;

[0018] S32. Based on the event flow model, extract sudden failure events of port machinery damage and equipment performance degradation, and construct TS fuzzy fault tree to identify port failure status from the perspectives of internal and external factors such as extreme environment, human error, and harsh working conditions.

[0019] S33. Transform the TS fuzzy fault tree into a directed acyclic graph of a Bayesian network, determine the Bayesian network model and node conditional probability table, wherein fuzzy numbers are used to describe the various fault states of nodes, and fuzzy subsets are used to characterize the probability of occurrence of each fault state.

[0020] S34. Use Bayes network bidirectional reasoning to solve the TS fuzzy fault tree, including two solution methods: pre-evaluation and dynamic evaluation. Obtain the fuzzy subset of the fault probability and the probability of fault occurrence for each state of the leaf node, and complete the fault state identification of the leaf node to determine the current state of the port operation system.

[0021] Furthermore, in step S32, constructing a TS fuzzy fault tree to identify the port fault status specifically includes the following sub-steps:

[0022] S321. Determine the top event, intermediate events, and basic events of the TS fuzzy fault tree;

[0023] S322. Determine the TS gate rules, i.e., the logical causal relationships between events;

[0024] S323. By associating the top event, intermediate events, and basic events through TS gate rules, an inverted tree diagram is constructed, which is the TS fuzzy fault tree;

[0025] Further, in step S33, converting the TS fuzzy fault tree into a directed acyclic graph of a Bayesian network, and determining the Bayesian network model and node conditional probability table specifically includes the following sub-steps:

[0026] S331. Determine the directed acyclic graph of the Bayesian network, that is, transform the bottom event, intermediate event, top event, and TS gate in the TS fuzzy fault tree into the root node, intermediate node, leaf node, and directed edge in the Bayesian network, respectively. Among them, the port production operation system fault ( T ) is the top event;

[0027] S332. Using the TS gate rule, construct the conditional probability table of Bayes network nodes and assign values ​​to them based on historical data to complete the establishment of the Bayes network model;

[0028] Furthermore, step S34 specifically includes the following sub-steps:

[0029] S341. Based on the analysis of historical port data and information, determine the fuzzy subset of failure probabilities for the root node (bottom event);

[0030] S342. For pre-assessment, it is used to predict the failure status of the port system before the occurrence of port emergencies; combining the Bayes network model and root node failure probability model obtained from S332 and S341, Bayes forward reasoning is performed to obtain the fuzzy subset of failure probabilities of intermediate nodes and leaf nodes in turn; in addition, the fuzzy importance and posterior probability of the root node can be calculated through Bayes backward reasoning to conduct emergency sensitivity analysis.

[0031] S343. For dynamic assessment, it is used for real-time fault status assessment of port systems after port emergencies; by utilizing fault alarm information from the CMS of crane management equipment and manual equipment inspection experience, combined with the Bayes network model and root node fault probability model obtained in S332 and S341, the fuzzy number of the current fault status of each root node is obtained, the membership degree of the fault status of each root node is calculated, and then the probability of occurrence of each fault status of the leaf nodes is calculated; in addition, the importance of the root node status can be obtained through Bayes backward reasoning, and sensitivity analysis can be performed.

[0032] S344. Based on the fuzzy subset of leaf node fault state probabilities or the probability of fault state occurrence obtained from pre-assessment or dynamic assessment, determine which of the three states the current port production operation system is in: normal, risk, or fault.

[0033] Furthermore, in step S4, the port self-healing fault-tolerant emergency control is carried out from both the system and equipment levels. The system-level self-healing fault-tolerant emergency control mainly addresses major port emergencies such as machine damage. Combining the two-layer programming concept, a port system self-healing fault-tolerant compensation optimization model is constructed. A system compensation optimization algorithm based on hyperheurism and potential field method is designed to solve the model, and finally, a system compensation optimization scheme is generated for the system to achieve online control driving for compensation optimization. The equipment-level self-healing fault tolerance mainly addresses common emergencies such as port equipment performance degradation. First, the probability interval of the performance parameters of risky equipment is predicted. Then, based on the capability prediction of risky equipment and the capability description of normal equipment, cooperative game decision-making for the performance optimization of multiple types of port equipment is carried out. With the goal of eliminating performance bottlenecks and reducing system losses, a port equipment performance optimization scheme is generated. Finally, behavior trees and situation diagrams are introduced to complete the mapping from the decision layer to the behavior layer, realizing online control of port equipment.

[0034] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0035] This invention designs a self-healing, fault-tolerant emergency control method for automated container ports. This method helps automated container ports achieve efficient online control of systems and equipment when faced with emergencies such as machine damage and equipment performance degradation. Through strategies such as compensatory optimization or risk prediction, it effectively reduces the impact of emergencies on subsequent port operations, greatly improves the emergency fault tolerance and risk control capabilities of automated container ports, and enhances the core competitiveness of ports. At the same time, it also provides a certain reference for the construction of self-healing, fault-tolerant control systems for other types of automated ports. Attached Figure Description

[0036] Figure 1 This is a flowchart of the automated container port self-healing fault-tolerant emergency control method of the present invention.

[0037] Figure 2 This is a schematic diagram of the automated container port fault status identification of the present invention.

[0038] Figure 3 Flowchart for generating port system compensation optimization scheme for this invention

[0039] Figure 4 Flowchart for generating port equipment performance optimization scheme for this invention Detailed Implementation

[0040] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0041] This invention proposes a self-healing fault-tolerant emergency control method for automated container ports. This method enables automated container ports to achieve efficient online control of systems and equipment in the face of emergencies such as machine damage and equipment performance degradation through compensatory optimization or risk prediction strategies. This effectively reduces the impact of emergencies on subsequent port operations, significantly improving the emergency fault tolerance and risk control capabilities of automated container ports, as well as their core competitiveness. Furthermore, it provides a reference for the construction of self-healing fault-tolerant control systems for other types of automated ports.

[0042] Specifically, it includes the following steps:

[0043] S1. Analyze the business relationships and technological paths of port production and operation, and construct a multi-priority task sequence for the current work shift at the port;

[0044] S2. Collect historical data from port operation systems and equipment to extract and classify the status characteristics of port emergencies.

[0045] S3. Construct a port state transition event flow model and a port system TS fuzzy fault tree to assess and identify the current port fault state;

[0046] S4. Based on the multi-priority task-driven approach of the work team, combined with the feedforward-feedback control concept, and based on the port status identification of the fault point and information on emergencies, carry out self-healing fault-tolerant emergency control at the system and equipment levels.

[0047] In the above examples, Figure 1 This document presents a flowchart of a self-healing and fault-tolerant emergency response method for automated container ports. The port emergencies described primarily refer to machine damage and equipment performance degradation. The former, caused by adverse operating conditions and factors, leads to major malfunctions, triggering subsequent events such as temporary production adjustments and exceeding limits. This requires the system to classify faulty equipment, restore its functionality, and conduct system integration testing. The latter, influenced by the natural environment, causes sensor malfunctions, affecting remote control and positioning, resulting in short-term equipment performance degradation and ultimately impacting the overall port performance. Therefore, an efficient and reasonable self-healing and fault-tolerant emergency response control method is needed to maintain the continuity of automated container port production operations. Specifically, port operations are conducted in shifts, each containing multiple tasks with different priorities, such as container loading and unloading, ship berth allocation, and AGV (Automated Guided Vehicle) operation. These tasks have certain priority relationships; therefore, to clearly describe the connections between tasks within each shift, it is necessary to analyze the current shift's situation. i By introducing the Directed Graph Theory (DRT), the current work shift is constructed. i A multi-priority task sequence.

[0048] In the above embodiments, the work shift plan mainly includes the ship operation plan, barge operation plan, equipment operation plan, and loading and unloading site plan; the process path refers to the process flow and execution sequence and path of related mechanical equipment adopted in the above business operations; the multi-priority task sequence of the current port work shift is constructed based on the Directed Task Graph (DRT), as shown in the following sub-steps:

[0049] S11. First, record the current work shift as... i , sorting out the work team i The relevant tasks and the process path for task execution;

[0050] S12. Secondly, construct the work team. i The set of directed graphs (DRTs) for real-time tasks, denoted as , where the directed graph for any task Characterized as , The set of job nodes representing a directed graph. The set representing the edges between job nodes;

[0051] S13. Subsequently, for the directed graph of the real-time task... Any job node Utilizing ordered pairs Characterize its longest execution time (WCET) and relative time limit (RD);

[0052] S14. Finally, for directed graphs of real-time tasks any edge in The minimum release interval parameter and relative time limit By comparing and controlling the order of tasks, a work shift can be formed. i A real-time multi-priority task sequence. When the work shift is completed... i After constructing the multi-priority task sequence, the key is how to conduct online control for emergencies occurring during work shift scheduling. Different emergencies have varying degrees of impact on the port system, requiring different control methods. Therefore, to achieve precise control, it is necessary to first extract features and classify common port emergencies. Specifically, historical data of port operation systems and equipment are obtained through the Terminal Operating System (TOS), Equipment Control System (ECS), and Crane Management System (CMS). Starting from dimensions such as emergency type, duration, and impact, principal component analysis (PCA) and K-means clustering analysis are comprehensively applied to classify and classify emergencies.

[0053] After classifying and grading port emergencies, it is necessary to identify the fault status of the current port operation system to determine whether it is in a risky or faulty state, in order to facilitate online control. The specific steps are as follows:

[0054] Step 1: First, based on historical failure case data collected from TOS, ECS and CMS, a port operation system event flow model is constructed by comprehensively using data cleaning and filtering, causal relationship analysis and other methods.

[0055] Step 2: Next, based on the event flow model, extract sudden failure events and construct a fuzzy fault tree for the port operation system TS from the perspectives of internal and external factors. Specifically:

[0056] 1. Determine the top event, intermediate events, and basic events of the TS fuzzy fault tree;

[0057] 2. Determine the TS gate rules, i.e., the logical causal relationships between events;

[0058] 3. By associating the top event, intermediate events, and basic events through TS gate rules, an inverted tree diagram is constructed, which is the TS fuzzy fault tree;

[0059] like Figure 2 As shown, where, x 1, x 2, x 3, x 4 is the bottom event. y 1, y 2 is an intermediate event. y 3 is the top event, which can also be denoted as T a, b, and c are TS fuzzy gates (defining the correlation between events). It is proposed to take port system risk failure as the top event, machine damage and equipment performance degradation as intermediate events, and severe working conditions, improper human operation, and poor natural environment as bottom events.

[0060] Step 3: Subsequently, to overcome the shortcomings of TS fuzzy fault tree computation being complex and unable to perform reverse reasoning, the TS fuzzy fault tree is transformed into a Bayesian network directed acyclic graph for efficient bidirectional reasoning. Through pre-assessment during work shifts and before the occurrence of emergencies, based on historical experience and expert inference, fuzzy subsets of the fault probability of each state of the Bayesian network leaf nodes are obtained. Through dynamic evaluation during work shifts and after unexpected events, based on the precise fault status of the root node or intermediate nodes. Determine the membership degree corresponding to each state of each root node. The probability of each fault state occurring at the leaf node is obtained. ;

[0061] Step 4: Finally, based on and Complete the fault status assessment and identification of leaf nodes (port operation system) under two modes. The specific steps are as follows:

[0062] 1. Based on the analysis of historical port data and information, determine the fuzzy subset of failure probabilities for the root node (bottom event);

[0063] 2. For pre-assessment, it is used to predict the failure status of the port system before the occurrence of port emergencies; combining the Bayes network model and root node failure probability model obtained from S332 and S341, Bayes forward reasoning is performed to obtain the fuzzy subsets of failure probabilities of intermediate nodes and leaf nodes in turn; in addition, the fuzzy importance and posterior probability of the root node can be calculated through Bayes backward reasoning to conduct emergency sensitivity analysis.

[0064] 3. For dynamic assessment, this method is used for real-time fault status assessment of port systems after port emergencies. Utilizing fault alarm information from the Crane Management System (CMS) and manual equipment inspection experience, combined with the Bayes network model and root node fault probability model obtained from S332 and S341, the fuzzy number of the current fault status of each root node is obtained. The membership degree of the fault status of each root node is calculated, and then the probability of each fault status occurring at the leaf nodes is calculated. Furthermore, the importance of the root node status can be obtained through Bayes backward reasoning for sensitivity analysis.

[0065] 4. Based on the fuzzy subset of leaf node fault state probabilities or fault state occurrence probabilities obtained from pre-assessment or dynamic assessment, determine which of the three states the current port production operation system is in: normal, risk, or fault.

[0066] After identifying the fault status of the port operation system for the current work shift, if the port is in a risky or faulty state, it is necessary to immediately analyze the type of the emergency (machine damage or equipment performance degradation) and carry out self-healing and fault-tolerant emergency control at the system or equipment level.

[0067] For major emergencies such as machine damage, it is necessary to implement system-level self-healing and fault-tolerant emergency control, such as... Figure 3 As shown. The specific steps are as follows:

[0068] Step 1: Obtain the system fault status and the current shift's work information and emergency information;

[0069] Step 2: Identify the problem domain of the port failure system;

[0070] Step 3: Construct a port system compensation optimization model, and carry out redundant equipment compensation planning and operation network reconfiguration optimization;

[0071] Step 4: Combining the compensatory optimization model and the Benders decomposition idea, design a compensatory optimization decision based on hyperheurism and potential field method. Use hyperheurism for population initialization and update iteration, and combine potential field method to complete the construction of port operation network, generating redundant compensatory optimization schemes for the main problem and system operation network reconstruction optimization schemes for sub-problems.

[0072] Step 5: Use the Benders concept to iteratively optimize the master-slave problem and generate the final compensatory optimization scheme for the system.

[0073] Step 6: Use AdaBoost and BPNN to obtain reasonable mapping decisions between port operation status and equipment selection rules and operation assignment rules;

[0074] Step 7: Finally, combining the compensation optimization scheme from Step 5 and the mapping decision knowledge from Step 6, the current shift's multi-priority task sequence is used to drive the compensation optimization scheme to complete the task allocation of the equipment, thereby achieving online scheduling and control.

[0075] In response to common emergencies such as performance degradation of port equipment, it is necessary to implement self-healing and fault-tolerant emergency control at the equipment level, such as... Figure 4 As shown, the specific steps are as follows:

[0076] Step 1: Identify any type of risky equipment i performance class and its conditional attribute set elements ;

[0077] Step 2: Construct the corresponding Naive Bayes classification model. Perform point prediction;

[0078] Step 3: Building The probability confidence interval;

[0079] Step 4: Construct port equipment types i Download any device j intelligent agents It conducts multi-agent cooperative game decision-making for various types of equipment to generate port equipment performance optimization solutions, among which... A Characteristic attributes (performance parameters, etc.) S Characterizes the state (idle, active, etc.). B Representing behavior, C Representation interaction, U Characterization of utility assessment (self-evaluation);

[0080] Step 5: Use Behavior Tree (BT) to define the port equipment operation process, such as behaviors like stopping, working (lifting, pitching, traveling), and standby.

[0081] Step 6: Construct a port equipment behavior decision-making layer model based on behavior trees to realize equipment behavior selection decisions;

[0082] Step 7: Construct a port equipment behavior layer model based on the situation map, and use the situation map to judge the port status and plan the operation path of the equipment.

[0083] Step 8: Send the equipment path planning scheme obtained in Step 7 to the ECS system in the form of instructions to realize automatic equipment scheduling.

[0084] The above embodiments are merely illustrative and explanatory of the present invention and are not intended to limit the invention to the scope of the described embodiments. Furthermore, those skilled in the art will understand that the present invention is not limited to the above embodiments, and many more variations and modifications can be made based on the teachings of the present invention, all of which fall within the scope of protection claimed by the present invention.

Claims

1. A self-healing fault-tolerant emergency control method for an automated container port, characterized in that, Comprise the following steps: S1. Comb the port production operation business association and its process path, build the multi-priority task sequence of the current shift of the port; the process path refers to the execution order and path of the process flow and related mechanical equipment adopted by the above business operation; the multi-priority task sequence of the current shift of the port is based on the real-time task directed graph DRT to build; S2. Collect the historical data of the port operation system and equipment, realize the state feature extraction and hierarchical classification of the port emergency; S3. Build a port state transition event stream model and a port system T-S fuzzy fault tree to assess and identify the fault state of the current port; the assessment and identification of the port fault state specifically comprises the following sub-steps: S31. Comb and build the event logic sequence and event stream model by using the data cleaning, data filtering preprocessing, and the theory and method of causal relationship analysis, combined with the historical data of the port system and equipment and the classification and grading of the emergency collected in step S2; S32. Based on the event stream model, extract the port machine damage and equipment performance degradation emergency, and from the extreme environment, human error, and adverse working conditions, build a T-S fuzzy fault tree for identifying the port fault state; S33. Convert the T-S fuzzy fault tree into a Bayes network directed acyclic graph to determine the Bayes network model and the node condition probability table, wherein the node's multiple fault states are described by using fuzzy numbers, and the occurrence probability of each fault state is depicted by using fuzzy subsets; S34. Solve the T-S fuzzy fault tree by using the Bayes network bidirectional reasoning, including two solving methods of pre-evaluation and dynamic evaluation, respectively obtain the fault probability fuzzy subset and the fault occurrence probability of each state of the leaf node, complete the fault state identification of the leaf node, and thus determine the state of the current port operation system; S4. Based on the multi-priority task driving of the shift, combined with the feedforward-feedback control idea, according to the identified state of the fault point and the emergency information, carry out self-healing fault-tolerant emergency regulation and control at the system and equipment levels; the port self-healing fault-tolerant emergency regulation is carried out at the system and equipment levels respectively; the self-healing fault-tolerant emergency regulation at the system level mainly deals with major emergencies of the machine-damaged port, combined with the double-layer planning idea, builds a port system self-healing fault-tolerant compensation optimization model, designs a system compensation optimization algorithm based on the hyper-heuristic and potential field method to solve the model, and finally generates a system compensation optimization scheme for online regulation and control driving of the system to realize compensation optimization; the self-healing fault-tolerant at the equipment level mainly deals with common emergencies of the performance degradation of the port equipment, first carries out the probability interval prediction of the risk equipment performance parameters, and then based on the ability prediction of the risk equipment and the ability description of the normal equipment, carries out the cooperative game decision of the performance optimization of the port multi-type equipment, generates the performance optimization scheme of the port equipment, and finally introduces the behavior tree and the situation map to complete the mapping from the decision layer to the behavior layer, and realizes the online regulation and control of the port equipment.

2. The self-healing fault-tolerant emergency regulation method for an automated container port according to claim 1, characterized in that, In the step S1, the port production operation business is carried out in units of shift operation plan, and the shift operation plan includes ship operation plan, vehicle barge operation plan, equipment operation plan, and loading and unloading site plan.

3. The self-healing fault-tolerant emergency control method for an automated container port according to claim 2, characterized in that, The step S1 is specifically shown in the following sub-steps: S11. First, the current shift is recorded as i , the relevant tasks of the shift i and the process path of task execution are combed; S12. Secondly, construct the work team. i The set of directed graphs (DRTs) for real-time tasks, denoted as , where the directed graph for any task Characterized as , The set of job nodes representing a directed graph. The set representing the edges between job nodes; S13. Subsequently, the directed graph for real-time tasks is updated any job node , using ordered pairs representing its worst-case execution time WCET and relative deadline RD; S14. Finally, the real-time task directed graph is ordered with respect to the minimum release interval time parameter and the relative deadlines of the tasks to form a real-time multi-priority task sequence for the shift. i S14. Finally, the real-time task directed graph is ordered with respect to the minimum release interval time parameter and the relative deadlines of the tasks to form a real-time multi-priority task sequence for the shift.​​​​​ 4. The self-healing fault-tolerant emergency regulation method for an automated container port according to claim 1, characterized in that, In the step S2, the port emergency event includes common machine damage and equipment performance degradation, the machine damage is mainly affected by severe working conditions and operation factors, and the equipment performance degradation is affected by natural environment; the historical data of the operation system and the equipment is collected through a terminal operating system (TOS), an equipment control system (ECS) and a crane management system (CMS); the extraction and classification of the state characteristics of the emergency event are based on the historical data collected by the system, the feature analysis and extraction of each dimension are performed from the dimensions of the emergency event type, the duration and the influence degree by using a principal component analysis (PCA) and a K-means clustering algorithm, and the classification of the emergency event is quickly realized, wherein the emergency event information is represented by a multi-dimensional feature sequence.

5. The self-healing fault-tolerant emergency control method for an automated container port according to claim 1, characterized in that, In the step S32, the T-S fuzzy fault tree is constructed to identify the port fault state, and the step S32 specifically includes the following sub-steps: S321. determining the top event, the intermediate event and the basic event of the T-S fuzzy fault tree; S322. determining the T-S gate rule, i.e. the logical causal relationship between events; S323. associating the top event, the intermediate event and the basic event by using the T-S gate rule to construct an inverted tree diagram, i.e. the T-S fuzzy fault tree.

6. The self-healing fault-tolerant emergency regulation method for an automated container port according to claim 1, characterized in that, In the step S33, the T-S fuzzy fault tree is converted into a directed acyclic graph of a Bayes network to determine the Bayes network model and the node condition probability table, and the step S33 specifically includes the following sub-steps: S331. determining the directed acyclic graph of the Bayes network, i.e. converting the bottom event, the intermediate event, the top event and the T-S gate in the T-S fuzzy fault tree into the root node, the intermediate node, the leaf node and the directed edge in the Bayes network, wherein the port production operation system fault T is the top event; S332. constructing the condition probability table of the Bayes network node by using the T-S gate rule and assigning values to the table according to historical data to complete the establishment of the Bayes network model.

7. The self-healing fault-tolerant emergency control method for an automated container port according to claim 1, characterized in that, In the step S34, the step S34 specifically includes the following sub-steps: S341. determining the fault probability fuzzy subset of the root node, i.e. the bottom event according to the historical data and the material analysis of the port; S342. for pre-evaluation, the step S342 is used for the port system fault state estimation before the occurrence of the port emergency event; the Bayes forward reasoning is performed by combining the Bayes network model and the root node fault probability model obtained in the steps S332 and S341 to sequentially obtain the fault probability fuzzy subset of the intermediate node and the leaf node; in addition, the root node fuzzy importance and the posterior probability can be calculated by the Bayes backward reasoning to perform the sensitivity analysis of the emergency event. S343. For dynamic evaluation, real-time fault state evaluation of port system after port emergency occurs; using crane management system (CMS) fault alarm information and device manual inspection experience, combining the Bayes network model and root node fault probability model obtained in S332 and S341, the current fault state fuzzy number of each root node is obtained, the fault state membership of each root node is calculated, and then the probability of occurrence of each fault state of the leaf node is calculated; in addition, the root node state importance can be obtained through Bayes reverse reasoning, and sensitivity analysis is performed; S344. According to the leaf node fault state probability fuzzy subset or fault state occurrence probability obtained by pre-evaluation or dynamic evaluation, it is judged which one of the three states of normal, risk or fault the current port production operation system is in.

Images