Flexible manufacturing key state estimation method based on asynchronous pulse nerve P system

Abstract

A flexible manufacturing key state estimation method based on an asynchronous pulse nerve P system relates to the technical field of intelligent manufacturing state estimation, and comprises the following steps: constructing a flexible manufacturing system model; defining a determined event rule set and a non-determined event rule set in the flexible manufacturing system model, and initializing state configuration; acquiring an observation sequence formed by a plurality of time sequence events in the flexible manufacturing system by using a sensor; dynamically calculating the excitation frequency upper bound of the event rule subset and the ground state configuration of the event rule subset based on the observation sequence by adopting an upper bound and ground state configuration calculation algorithm; constructing an excitation frequency constraint of the non-deterministic event rule; constructing a state evaluation formula according to the excitation frequency constraint of the non-deterministic event rule, the excitation frequency of the non-deterministic event rule, the basic state configuration of the non-deterministic event and the flexible manufacturing system model; estimating a state set of the flexible manufacturing system by adopting a state evaluation formula; the method is used for solving the problem of low accuracy of traditional state estimation.

Description

Technical Field

[0001] This invention relates to the field of intelligent manufacturing state estimation technology, specifically to a method for estimating critical states in flexible manufacturing based on an asynchronous pulse neural P system. Background Technology

[0002] In modern manufacturing, flexible manufacturing systems (FMS) have gradually become an important direction for industrial development due to their ability to quickly allocate resources to adapt to changing market demands. This rapid response capability not only significantly improves production efficiency but also better supports the personalized and diversified production of products. However, in actual operation, due to the complexity of the production environment, the diversity of processes, and the influence of external uncertainties, how to effectively monitor and control the manufacturing process has become a major challenge, especially under the condition of limited sensor configuration, it is particularly difficult to accurately obtain system status information.

[0003] To address the challenges posed by the complex production environment, state estimation is currently employed to solve related problems. In flexible manufacturing scenarios, frequent uncertain events, such as equipment failures, unstable material supply, and temporary order adjustments, can disrupt normal production. Therefore, the system needs to possess strong dynamic response capabilities to adjust production strategies in a timely manner and accurately assess and predict critical states. Thus, effective state estimation provides crucial information for decision-making, whether for production scheduling or real-time feedback on the production process, helping to improve resource utilization efficiency and achieve higher economic benefits.

[0004] However, traditional state estimation methods generally use probabilistic models and static analysis. Probabilistic models usually rely on idealized assumptions such as "events are independent" or "state transition probabilities are fixed". They are good at handling random events with known probability distributions, but have low accuracy in judging uncertain events, thus affecting the state estimation results of flexible manufacturing systems. In addition, static analysis requires batch processing of large amounts of data, which is time-consuming. Therefore, it is slow to respond to sudden events, which can lead to delayed decision-making, missing the best adjustment window, and causing production blockage or waste of resources.

[0005] Chinese patent CN117852825A discloses a deadlock-free scheduling method for a flexible manufacturing system with central resources based on deep learning. This method provides a two-step look-forward deadlock avoidance strategy to avoid deadlock states during scheduling. In addition, the method also uses an improved Dijkstra algorithm based on a feedforward neural network to determine the processing time of the production state and select the optimal production state for expansion, which can quickly find a scheduling sequence that meets the requirements and improve production efficiency. However, this method focuses more on the planning of scheduling paths and deadlock prevention, and has a weak ability to judge uncertain events, resulting in low accuracy of state estimation.

[0006] Therefore, we propose a method to improve the accuracy of state estimation. Summary of the Invention

[0007] The purpose of this invention is to provide a method for estimating critical states in flexible manufacturing based on an asynchronous pulse neural P system, which addresses the problem of low accuracy in traditional state estimation.

[0008] This invention is achieved through the following technical solution: A method for estimating critical states in flexible manufacturing based on an asynchronous pulse neural P system, specifically including: Based on vector pulse neural The system constructs a flexible manufacturing system model; Define the set of deterministic event rules and the set of non-deterministic event rules in the flexible manufacturing system model, and initialize the state configuration; Utilize sensors to acquire observation sequences consisting of multiple time-series events in a flexible manufacturing system; An upper bound and ground state configuration calculation algorithm is adopted to dynamically calculate the upper bound of the excitation number of the event rule subset and the ground state configuration of the event rule subset based on the observation sequence; Construct constraints on the number of times nondeterministic event rules are triggered; Based on the constraints of the number of times the nondeterministic event rule is triggered, the number of times the nondeterministic event rule is triggered, the basic state configuration of the nondeterministic event, and the flexible manufacturing system model, a state evaluation formula is constructed. The state set of the flexible manufacturing system is estimated using a state evaluation formula.

[0009] Furthermore, the algorithm for calculating the upper bound and ground state configuration, based on the observation sequence, dynamically calculates the upper bound of the excitation count of the event rule subset and the ground state configuration of the event rule subset, specifically including: Initialize the calculation parameters; Obtain any event in the observation sequence ; If the event To determine an event, a subset of the rules for that event is extracted. Based on the conflict relationship between the subset of event rules and the non-deterministic event rules, the upper bound of the excitation count of the subset of event rules and the ground state configuration of the subset of event rules are dynamically calculated. If the event If the event is non-deterministic, then directly update the event rule set. The upper bound of the number of excitations, and the ground state configuration remains unchanged, i.e. .

[0010] Furthermore, the initialization calculation parameters include: observation sequence Set to empty, that is ; Ground state configuration of the flexible manufacturing system model Set as the initial system configuration ,Right now ; For each nondeterministic event Its rule subset The upper bound for the number of excitations is set to 0. and its set of rules The upper bound for the number of excitations is set to 0, that is... .

[0011] Furthermore, the aforementioned if event To determine an event, a subset of rules for that event is extracted. Based on the conflict relationship between this subset of rules and the rules for non-deterministic events, the upper bound of the excitation count of the subset of rules and the ground state configuration of the subset of rules are dynamically calculated, specifically including: If the event rules are a subset It does not conflict with any non-deterministic rule, that is:

[0012] Then update the base state configuration:

[0013] in This indicates the impact of the rule's execution on the system state; If the event rules are a subset The inputs and outputs that affect nondeterministic rules are:

[0014] Then the activation probability of the nondeterministic rule is recalculated and constrained; like If so, the ground state configuration is updated directly, and the upper bound of the excitation count of the nondeterministic rule for that event is adjusted synchronously.

[0015] Furthermore, the subset of the event rules... The inputs and outputs that affect nondeterministic rules, i.e. Then, the activation probability of the nondeterministic rule is recalculated and constrained, specifically including: Set vector The vector dimension is ; Find the subset of rules for this event Related set of nondeterministic rules ; subset of rules Each rule Based on the current basic observation status Calculate its excitation number ; For each rule subset ,if If so, then update the upper bound of the number of triggers; Finally, update the observation sequence. New events were subsequently observed. The ground state configuration.

[0016] Furthermore, the number of times it is stimulated The calculation formula is:

[0017] In the formula, This indicates that if the activation rule is Then the neuron Consumption pulse resources , This indicates that if the activation rule is Then the neuron Will receive pulse resources , Represents neurons The number of pulses; For each rule subset ,if Then, update the upper bound of the number of excitations, and calculate it using the following formula:

[0018] In the formula, Subset of rules The upper bound of the number of activations of the rule.

[0019] Furthermore, the update is in the observation sequence New events were subsequently observed. The ground state configuration is calculated using the following formula:

[0020] In the formula, This is the synaptic rule relation matrix. Represents a sequence The effective rule triggering mode.

[0021] Furthermore, the aforementioned if event For non-deterministic events, the upper bound of the trigger count of the event's rule set is directly updated, specifically including: For all satisfying For a subset of rules, first update the upper bound of the number of triggers:

[0022] In the formula, Subset of rules All rules in the system state The sum of enabling degrees; Then update the rule set. Upper bound of the number of excitations: .

[0023] Furthermore, the specific formula for constraining the number of triggers in constructing the nondeterministic event rule is as follows: .

[0024] Furthermore, the construction state evaluation formula is specifically as follows: .

[0025] The technical solution of the present invention has at least the following advantages and beneficial effects: This invention discloses a method for estimating critical states in flexible manufacturing based on an asynchronous pulse neural P system. By explicitly defining a set of rules for nondeterministic events and constructing mathematical constraints on their excitation frequency, the method transforms uncertainties that are difficult to quantify into computable constraint problems. This enables the method to more accurately infer the true resource distribution and critical states within the system under conditions of limited sensors.

[0026] In addition, by using the upper bound and ground state configuration calculation algorithm, the state estimate can be dynamically and recursively updated based on the real-time acquired time-series event observation sequence. This method enables rapid tracking and response to changes in system state, avoids production blockage or resource waste caused by decision lag, and provides a reliable basis for real-time optimization scheduling of production.

[0027] Furthermore, by distinguishing between deterministic and non-deterministic events and employing different ground state configuration updates and upper bound calculation strategies for excitation counts, the algorithm logic becomes clearer, computational resource allocation becomes more efficient, and the robustness and practicality of the method in handling mixed event scenarios are enhanced. Attached Figure Description

[0028] Figure 1 This is a schematic diagram of a method for estimating critical states in flexible manufacturing based on an asynchronous pulse neural P system according to the present invention. Figure 2 This is a structural diagram of an example of the present invention; Figure 3 This is a schematic diagram of the electronic device in this invention. Detailed Implementation

[0029] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0030] Example 1 like Figure 1 The method for estimating critical states in flexible manufacturing based on an asynchronous pulse neural P system, as shown, specifically includes: Based on vector pulse neural The system constructs a flexible manufacturing system model; In particular, flexible manufacturing systems ,in This is the synaptic rule relation matrix of the system. This is the initial state of the system. The system uses a pulse alphabet set and also includes a labeling function. The label function is used to map rules to a set of events. Furthermore, this flexible manufacturing system meets the following four technical conditions: (A1) The system structure is known, meaning that the system's neurons, rules, and their interconnections are clearly defined. This is the foundation for modeling, requiring that the system's "skeleton" is clear, i.e., all working components (neurons), their behavioral rules (activation rules), and their interconnections (synapses) are clearly defined. If the structure is unknown, the entire model cannot be established, and subsequent state estimation becomes impossible. (A2) Initial state Given that, it represents the system at time. The distribution of resources in each neuron at any given time is the starting point for state evolution. It requires that the number of resources (pulse distribution) in each working component be known at the beginning of estimation, providing a reliable and unique initial reference point for subsequent state deduction based on event observations. If the starting point is unknown, any subsequent estimation will lose its benchmark. (A3) It has sensors or hardware devices that can observe regular triggering events, which are the input source for the algorithm to run; it requires the system to provide external, objective measurement data—that is, the "time-series event observation sequence". This observation sequence is the "fuel" that drives your entire dynamic estimation algorithm. Without this input, the algorithm cannot perceive changes in the system and therefore cannot update its state. (A4) Nondeterministic rules are non-contact rules, meaning that excitations between different rules do not affect each other, satisfying the following formula:

[0031] This is a key constraint to ensure the computability of the method.

[0032] Define the set of deterministic event rules and the set of non-deterministic event rules in the flexible manufacturing system model, and initialize the state configuration; The purpose of this step is to structurally model the uncertainties of the system using prior knowledge and to provide a definite initial reference point for the dynamic process. Specifically, this includes: By defining a set of rules for deterministic events and a set of rules for indeterminate events, a rule-based foundation can be provided for subsequent analysis of specific events. It should be noted that the set of rules for deterministic events defines all production behaviors in a flexible manufacturing system that can be directly and clearly observed, while the set of rules for indeterminate events defines all production behaviors that exist in the system but cannot be directly observed. Utilize sensors to acquire observation sequences consisting of multiple time-series events in a flexible manufacturing system; When a flexible manufacturing system is in operation, it may generate a variety of events, such as loading, processing, internal transfer, and quality inspection. Loading, processing, and quality inspection are observable, i.e., deterministic events, while internal transfer is difficult to observe directly, so it is a non-deterministic event. Therefore, the set of all events that occur along the time axis constitutes the observation sequence.

[0033] An upper bound and ground state configuration calculation algorithm is adopted to dynamically calculate the upper bound of the excitation number of the event rule subset and the ground state configuration of the event rule subset based on the observation sequence; This step uses a dynamic computation mechanism with a "dual track" to separate and coordinate the determinism and uncertainty of the system under partial observation. The ground state configuration represents the state evolution path of the system under the "most economical" or "most ideal" condition. It assumes that only deterministic events that are clearly observed occur and updates the system state strictly in the order of their observation. Thus, its calculation and update provide a deterministic and precisely calculable state evolution baseline. The upper bound of the number of times the event rule subset is triggered represents the maximum number of times these invisible rules can be triggered under the current observation and resource constraints. The mathematical constraints quantify the magnitude of uncertainty and reflect the hard constraints of the system's physical resources on hidden activities.

[0034] Furthermore, the algorithm for calculating the upper bound and ground state configuration, based on the observation sequence, dynamically calculates the upper bound of the excitation count of the event rule subset and the ground state configuration of the event rule subset, specifically including: Initialize the calculation parameters, including: observation sequence Set to empty, that is ; Ground state configuration of the flexible manufacturing system model Set as the initial system configuration ,Right now ; For each nondeterministic event Its rule subset The upper bound for the number of excitations is set to 0. and its set of rules The upper bound for the number of excitations is set to 0, that is... ; Obtain any event in the observation sequence ; If the event To determine an event, a subset of the rules for that event is extracted. Based on the conflict relationship between the subset of event rules and the non-deterministic event rules, the upper bound of the excitation count of the subset of event rules and the ground state configuration of the subset of event rules are dynamically calculated. If the event If the event is non-deterministic, then directly update the event rule set. The upper bound of the number of excitations specifically includes: For all satisfying For a subset of rules, first update the upper bound of the number of triggers:

[0035] In the formula, Subset of rules All rules in the system state The sum of enabling degrees; Then update the rule set. Upper bound of the number of excitations: ; The ground state configuration remains unchanged, that is... .

[0036] Furthermore, the specific steps for dynamically calculating the upper bound of the excitation count of the event rule subset and the ground state configuration of the event rule subset include: If the event rules are a subset It does not conflict with any non-deterministic rule, that is:

[0037] In the formula, These are mapping operation functions, used to perform a class of mapping operations from rules to sets of neurons; Indicates by stimulating the set Each rule in The union of subsets of neurons that can produce pulse output; Then update the base state configuration. ,in This indicates the impact of the rule's execution on the system state; If the event rules are a subset The inputs and outputs that affect nondeterministic rules are:

[0038] Recalculate and constrain the excitation probability of nondeterministic rules; like If so, the ground state configuration is updated directly, and the upper bound of the excitation count of the nondeterministic rule for that event is adjusted synchronously.

[0039] As needed, the excitation probability of nondeterministic rules is recalculated and constrained. Specific steps include: Set vector The vector dimension is ; Find the subset of rules for this event Related set of nondeterministic rules ; In the formula, For neurons The input rule set, and To activate A set of neurons that consumes resource 'a' The intersection of the set of neurons that, when activated by nondeterministic rules, increase resource a; subset of rules Each rule Based on the current basic observation status Calculate its excitation number The calculation formula is:

[0040] In the formula, This indicates that if the activation rule is Then the neuron Consumption pulse resources , This indicates that if the activation rule is Then the neuron Will receive pulse resources , Represents neurons The number of pulses; For each rule subset ,if Then update the upper bound of the number of triggers. The calculation formula is:

[0041] In the formula, Subset of rules The upper bound of the number of activations in the middle rule; Finally, update the observation sequence. New events were subsequently observed. ground state configuration The calculation formula is:

[0042] In the formula, This is the synaptic rule relation matrix. Represents a sequence The effective rule triggering mode.

[0043] Constructing the trigger count constraint for nondeterministic event rules ; The specific formula is as follows:

[0044] in, The elements in the set must satisfy the following condition: for any subset of rules, the sum of its excitation counts cannot exceed the upper bound. and rule set The sum of the activation counts of all rules equals the upper bound. ; The principle of this step is based on the conservation of physical resources, establishing strict mathematical boundaries for unobservable system behavior, thereby compressing infinite possibilities into a finite, computable solution space. Its fundamental principle stems from a basic fact: in a flexible manufacturing system, any production action (the activation of rules) must consume and generate a specific amount of resources.

[0045] Based on the constraints of the number of triggering events of the nondeterministic event rules, the number of triggering events of the nondeterministic event rules, the basic state configuration of the nondeterministic events, and the flexible manufacturing system model, a state evaluation formula is constructed. The specific formula is as follows:

[0046] in, This is the ground-state configuration for flexible manufacturing systems. For the synaptic matrix of nondeterministic rules in a flexible manufacturing system, The number of times the rule is triggered is a vector. In relation to the event The relevant set of nondeterministic rule execution vectors; additionally, in flexible manufacturing systems, the state... This indicates the amount of resources remaining in each working component (neuron) at the current moment. This is crucial for estimating the state of each stage in the production process and can reflect whether the resource distribution is reasonable and whether there are bottlenecks. ground state configuration The system represents the observed event sequence The initial state of the system is the most basic state of the system without considering the number of times nondeterministic rules are triggered. It is the basis for the evolution of subsequent states and determines the range of states that the system may enter. Upper bound on the number of activations Used to restrict a subset of rules The maximum number of times each rule is triggered corresponds to the maximum number of times certain work components (such as production equipment) can perform certain operations under resource constraints in a flexible manufacturing system; for example, the maximum number of workpieces a workstation can process in its current state. Nondeterministic rule execution vector set Indicates that when an event is observed Subsequently, which nondeterministic rules might be activated and the number of times they are activated reflect the state changes of working components in the system where uncertainty exists (such as production areas lacking sensor monitoring). State estimation formula It describes how to infer the possible states of a system from observed sequences of events; for flexible manufacturing systems, this means being able to infer the resource status of all workstations on the production line based on limited sensor data and identify potential anomalies or malfunctions. The state set of the flexible manufacturing system is estimated using a state evaluation formula.

[0047] It should be noted that the method in this embodiment is particularly suitable for complex manufacturing environments with limited sensing devices and uncertain events. This method can achieve efficient estimation of critical states by effectively utilizing limited observation data. Thus, when the number of sensing devices is insufficient, the algorithm can dynamically integrate multi-source information through the adaptive characteristics of spiking neural networks and respond to changes in system state in a timely manner, thereby improving the accuracy and reliability of state estimation.

[0048] Furthermore, this state estimation method can flexibly address the challenges posed by uncertainty in environments with non-deterministic events. By employing asynchronous information processing and a multi-level state information exchange mechanism, it can still make reasonable inferences and predictions based on existing state information, even when specific events cannot be fully observed. This method effectively enhances the monitoring capability of critical states, ensuring that flexible manufacturing systems can continuously optimize production processes and improve work efficiency under the constraints of sensing devices and environmental uncertainties, thereby providing solid technical support for intelligent manufacturing.

[0049] Furthermore, a table summarizing the meanings of the symbols in the formula of Example 1 is provided below:

[0050] Example 2 As an example, we describe the state estimation of a flexible manufacturing system as a problem of determining the current state of the system. We formalize this using previously defined notation as follows: 1. System Structure Consider a pulsatile nerve system ,like Figure 2 As shown, and 'a' in the figure represents a pulse. The connection lines between neurons can be understood as interfaces, connectors, or adapters between working components (neurons), used to enable the transfer and interaction of information and resources between different working components; To trigger rules for the system.

[0051] In this system, nondeterministic rule set Including rules Deterministic rule set Including rules The initial state of the system is:

[0052] 2. Beginning of the observation process: Initially, the observation device does not observe any rule execution. Therefore, the state set can be determined as follows:

[0053] 3. Observation of rule execution: Hypothetical events It was observed that three nondeterministic rules that could be executed resulted in the state set being updated as follows:

[0054] 4. Accumulation of multiple observations: If the observed event It happened again, that is Then the rules can be determined. and It can be executed at most once, and the rule It may also be executed twice. Therefore, the updated state set is:

[0055] 5. The impact of deterministic rules: If deterministic rules... Execution allows us to infer the previous observation. Not derived from rules The result of executing the command twice is used to determine the rule. Released, execution is possible, state set updated to:

[0056] 6. Third observation: Hypothetical event The third observation, based on the previous state. The possible sequence of rules to be executed is as follows: , Therefore, the state set can be updated as follows:

[0057] 7. Final observation conclusion: If deterministic rules If observed, then we can conclude that only the set of states exists. The first state in the sequence is compatible with the last observation, therefore the result is:

[0058] The final confirmed sequence of rules executed is as follows:

[0059] That is, in flexible manufacturing systems where hardware is constrained (and there are nondeterministic observations) In the case of natural persons, even a natural person can still estimate a definite sequence of events based on observational data. This patented technology, however, does not rely on deduction by a natural person, but solely on data based on pulse nerves. The system's computational process solves the state estimation problem.

[0060] This example also discusses how to integrate spiking neural networks. The system state estimation method is applied to the state assessment problem of flexible manufacturing systems. The specific application steps are as follows: S1 System Configuration and Initialization: Considering a spiking nerve system ,like Figure 2 As shown. Its rule set The set of nondeterministic rules Deterministic rule set Define the initial state as:

[0061] The nondeterministic synaptic rule relation matrix is ​​as follows:

[0062] S2 Observation Events and Status Updates: In the absence of observed events, the baseline configuration is as follows: And for all ,have (in When new events are observed At that time, according to the upper bound and ground state configuration Calculation algorithm rules and In initial configuration The following is enabled. To simplify the description, we will use subsets. Represent as Therefore, we update:

[0063] The value is 1, and the ground state remains unchanged, as shown in the second row of Table 1. Therefore

[0064] Using this data, we calculate the state set based on the nondeterministic rule excitation frequency constraint and the state evaluation formula. First, based on the nondeterministic rule excitation frequency constraint:

[0065] By analogy, we can conclude that:

[0066] Secondly, based on the state evaluation formula, the update of the state set is directly calculated:

[0067] S3 Multiple Observations and Subsequent Updates In observation At that time, continue calling the upper bound. and ground state configuration The calculation algorithm will continue to be updated.

[0068] It is 2, and the ground state The values ​​remain unchanged, as shown in the third row of Table 1. The situation.

[0069] If the observation is We call the upper bound. and ground state configuration The computational algorithm, due to Satisfy the upper bound and ground state configuration computational algorithms This situation; Calculation yielded:

[0070] The status is then updated as follows:

[0071] Thus update all containing of The upper realm Specifically:

[0072] Finally, the basic state is obtained:

[0073] Iterative calculations can obtain the basic state for each estimation. As shown in List 1 below: Table 1. Basic State Diagram of Event Estimation

[0074] Finally, it can be observed that Calculation obtained According to the state estimation formula We can obtain:

[0075] The final confirmed sequence of rules executed is as follows:

[0076] Through the steps described above, this example demonstrates a pulse neural network-based approach. The specific implementation process of the system's flexible manufacturing system state estimation method includes system initialization, state observation and updating, and subsequent configuration adjustments resulting from multiple observations.

[0077] Example 3 As attached Figure 3 An electronic device shown includes: Processor, memory, communication interface; The memory is used to store the executable instructions of the processor; The processor is configured to execute the aforementioned flexible manufacturing critical state estimation method based on an asynchronous spiking neural P system by executing the executable instructions.

[0078] A readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-described method for estimating critical states in flexible manufacturing based on an asynchronous spiking neural P system.

[0079] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for estimating critical states in flexible manufacturing based on an asynchronous impulse neural P system, characterized in that, Specifically, it includes: Based on vector pulse neural The system constructs a flexible manufacturing system model; Define the set of deterministic event rules and the set of non-deterministic event rules in the flexible manufacturing system model, and initialize the state configuration; Utilize sensors to acquire observation sequences consisting of multiple time-series events in a flexible manufacturing system; An upper bound and ground state configuration calculation algorithm is adopted to dynamically calculate the upper bound of the excitation number of the event rule subset and the ground state configuration of the event rule subset based on the observation sequence; Construct constraints on the number of times nondeterministic event rules are triggered; Based on the constraints of the number of times the nondeterministic event rule is triggered, the number of times the nondeterministic event rule is triggered, the basic state configuration of the nondeterministic event, and the flexible manufacturing system model, a state evaluation formula is constructed. The state set of the flexible manufacturing system is estimated using a state evaluation formula.

2. The method for estimating critical states in flexible manufacturing based on an asynchronous pulse neural P system according to claim 1, characterized in that: The algorithm employing upper bounds and ground state configuration calculations dynamically calculates the upper bound of the excitation count of the event rule subset and the ground state configuration of the event rule subset based on the observation sequence, specifically including: Initialize the calculation parameters; Obtain any event in the observation sequence ; If the event To determine an event, a subset of the rules for that event is extracted. Based on the conflict relationship between the subset of event rules and the non-deterministic event rules, the upper bound of the excitation count of the subset of event rules and the ground state configuration of the subset of event rules are dynamically calculated. If the event If the event is non-deterministic, then directly update the event rule set. The upper bound of the number of excitations, and the ground state configuration remains unchanged, i.e. .

3. The method for estimating critical states in flexible manufacturing based on an asynchronous pulse neural P system according to claim 2, characterized in that: The initialization calculation parameters include: observation sequence Set to empty, that is ; Ground state configuration of the flexible manufacturing system model Set as the initial system configuration ,Right now ; For each nondeterministic event Its rule subset The upper bound for the number of excitations is set to 0. and its set of rules The upper bound for the number of excitations is set to 0, that is... .

4. The method for estimating critical states in flexible manufacturing based on an asynchronous pulse neural P system according to claim 3, characterized in that: If the event To determine an event, a subset of rules for that event is extracted. Based on the conflict relationship between this subset of rules and the rules for non-deterministic events, the upper bound of the excitation count of the subset of rules and the ground state configuration of the subset of rules are dynamically calculated, specifically including: If the event rules are a subset It does not conflict with any non-deterministic rule, that is: In the formula, These are mapping operation functions, used to perform a class of mapping operations from rules to sets of neurons; Indicates by stimulating the set Each rule in The union of subsets of neurons that can produce pulse output; Then update the base state configuration: in This indicates the impact of the rule's execution on the system state; If the event rules are a subset The inputs and outputs that affect nondeterministic rules are: Then the activation probability of the nondeterministic rule is recalculated and constrained; like If so, the ground state configuration is updated directly, and the upper bound of the excitation count of the nondeterministic rule for that event is adjusted synchronously.

5. The method for estimating critical states in flexible manufacturing based on an asynchronous pulse neural P system according to claim 4, characterized in that: If the event rule subset The inputs and outputs that affect nondeterministic rules, i.e. Then, the activation probability of the nondeterministic rule is recalculated and constrained, specifically including: Set vector The vector dimension is ; Find the subset of rules for this event The relevant set of nondeterministic rules: In the formula, For neurons The input rule set, and To activate A set of neurons that consumes resource 'a' The intersection of the set of neurons that, when activated by nondeterministic rules, increase resource a; subset of rules Each rule Based on the current basic observation status Calculate its excitation number ; For each rule subset ,if If so, then update the upper bound of the number of triggers; Finally, update the observation sequence. New events were subsequently observed. The ground state configuration.

6. The method for estimating critical states in flexible manufacturing based on an asynchronous pulse neural P system according to claim 5, characterized in that: Number of activations The calculation formula is: In the formula, This indicates that if the activation rule is Then the neuron Consumption pulse resources , This indicates that if the activation rule is Then the neuron Will receive pulse resources , Represents neurons The number of pulses; For each rule subset ,if Then, update the upper bound of the number of excitations, and calculate it using the following formula: In the formula, Subset of rules The upper bound of the number of activations of the rule.

7. The method for estimating critical states in flexible manufacturing based on an asynchronous pulse neural P system according to claim 6, characterized in that: The update is in the observation sequence New events were subsequently observed. The ground state configuration is calculated using the following formula: In the formula, This is the synaptic rule relation matrix. Represents a sequence The effective rule triggering mode.

8. The method for estimating critical states in flexible manufacturing based on an asynchronous pulse neural P system according to claim 7, characterized in that: If the event For non-deterministic events, the upper bound of the trigger count of the event's rule set is directly updated, specifically including: For all First, update the upper bound of the number of triggers: In the formula, Subset of rules All rules in the system state The sum of enabling degrees; Then update the rule set. Upper bound of the number of excitations: 。 9. The method for estimating critical states in flexible manufacturing based on an asynchronous pulse neural P system according to claim 8, characterized in that: The specific formula for constraining the number of triggers in constructing the nondeterministic event rule is as follows: 。 10. The method for estimating critical states in flexible manufacturing based on an asynchronous pulse neural P system according to claim 9, characterized in that: The construction status evaluation formula is as follows: 。

Images