Initial resource allocation method, device and equipment for automatic manufacturing system

Abstract

The invention discloses an initial resource allocation method, device and equipment of an automatic manufacturing system, relates to the field of automation, and is used for balancing minimum initial resource allocation and operation robustness of the automatic manufacturing system. The method comprises the following steps: carrying out LPN modeling on an automatic manufacturing system to obtain a label Petri net; if the tag Petri net only contains considerable transition, estimating a first minimum initial identifier and a second minimum initial identifier in the tag Petri net according to a tag sequence indicating the target task / process; if the label Petri network contains the unobservable transition, estimating a first minimum initial identifier in the label Petri network according to the label sequence; and finally, allocating the resource indicated by the obtained minimum initial identifier to the target task / process. The method can be applied to the LPN models only containing the considerable transition and the LPN models containing the inconsiderable transition at the same time, and stable and minimum resource allocation of the system is achieved.

Description

Technical Field

[0001] The present invention relates to the field of automation technology, and in particular to an initial resource allocation method, device and equipment for an automatic manufacturing system. Background Art

[0002] Over the past few decades, with the widespread application of information technology, automation technology, and computing technology, traditional manufacturing systems have gradually transformed into automated manufacturing systems, with the goal of significantly reducing manufacturing costs, improving product quality, ensuring production safety, and being able to quickly respond to market changes and customization requirements.

[0003] In practical production applications, initial resource allocation in automated manufacturing systems—that is, allocating the minimum number of resources while meeting predetermined tasks, thereby achieving cost savings and maximizing economic benefits—is of great practical significance. Inspired by this initial resource allocation problem, the use of labeled Petri nets to estimate minimum initial identities has become a research hotspot. In a labeled Petri net, an observed label sequence represents a given target task sequence or process sequence. A minimum initial identity is one that, given the Petri net structure, can trigger at least one transition sequence consistent with (corresponding to) the label sequence and contains the fewest total tokens. In other words, the minimum initial identity is the one with the minimum number of tokens (representing resources), calculated from the observed label sequence and representing the minimum number of resources.

[0004] Currently, almost all research focuses on minimizing resource consumption in automated manufacturing systems. However, the fundamental task of automated manufacturing systems is to produce and process products to deliver them on schedule. Automated manufacturing systems are constantly exposed to external uncertainties, and their operational robustness warrants attention. Robustness refers to a system's ability to maintain its performance and functionality in the face of various types of perturbations, changes, or uncertainties. Summary of the Invention

[0005] The object of the present invention is to provide an initial resource allocation method, device and equipment for an automatic manufacturing system to address all or part of the above-mentioned problems, so as to balance the minimum initial resource allocation and operational robustness of the automatic manufacturing system.

[0006] The technical solution adopted in the present invention is as follows:

[0007] An initial resource allocation method for an automatic manufacturing system, comprising:

[0008] S1. Perform LPN modeling on the automatic manufacturing system to obtain a labeled Petri net; if the labeled Petri net contains only observable transitions, execute step S2; if the labeled Petri net contains unobservable transitions, execute step S3;

[0009] S2. Estimate a first minimum initial identifier and a second minimum initial identifier in a label Petri net based on a label sequence indicating a target task / process; the first minimum initial identifier enables at least one transition sequence corresponding to the label sequence, and the second minimum initial identifier enables all transition sequences corresponding to the label sequence;

[0010] S3. Estimating a first minimum initial identifier in a label Petri net according to the label sequence;

[0011] S4. Allocate the resource indicated by the obtained minimum initial identifier to the target task / process.

[0012] In another aspect, the present application further provides an initial resource allocation device for an automatic manufacturing system, comprising:

[0013] The first module is used to perform LPN modeling on the automatic manufacturing system to obtain a labeled Petri net; if the labeled Petri net contains only observable transitions, the second module is activated; if the labeled Petri net contains unobservable transitions, the third module is activated;

[0014] The second module is configured to estimate a first minimum initial identifier and a second minimum initial identifier in a label Petri net based on a label sequence indicating a target task / process; the first minimum initial identifier enables at least one transition sequence corresponding to the label sequence, and the second minimum initial identifier enables all transition sequences corresponding to the label sequence;

[0015] A third module is configured to estimate a first minimum initial identifier in a label Petri net according to the label sequence;

[0016] The fourth module allocates the resources indicated by the obtained minimum initial identifier to the target task / process.

[0017] In addition, the present application also provides an initial resource allocation device for an automatic manufacturing system, which includes a processor and a storage medium, wherein the storage medium stores computer instructions. When the processor runs the computer instructions, it can execute the above-mentioned initial resource allocation method for the automatic manufacturing system.

[0018] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0019] The initial resource allocation scheme for the automatic manufacturing system proposed in this application can simultaneously solve the minimum initial identification problem for the LPN model abstracted by the automatic manufacturing system that only contains observable changes and the LPN model that contains unobservable changes, thereby realizing the allocation of minimum initial resources; and while performing the minimum initial resource allocation, the robustness of the automatic manufacturing system is also taken into account, thereby realizing robust minimum resource allocation for the automatic manufacturing system. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The present invention will now be described by way of example with reference to the accompanying drawings, in which:

[0021] Figure 1 This is a flow chart of the initial resource allocation method for the automatic manufacturing system provided in an embodiment of the present application.

[0022] Figure 2 It is a network model diagram of a labeled Petri net containing only observable transitions modeled in a specific application of an embodiment of the present application.

[0023] Figure 3 corresponds to Figure 2 The structure diagram of the label synthesis network.

[0024] Figure 4 Yes Figure 2 Schematic diagram of a transition sequence for estimating a second minimum initial identifier according to an embodiment.

[0025] Figure 5 Yes Figure 2 Schematic diagram of a transition sequence for estimating a first minimum initial identifier in an embodiment.

[0026] Figure 6 It is a network model diagram of a labeled Petri net containing unobservable transitions modeled in a specific application of an embodiment of the present application.

[0027] Figure 7 Yes Figure 6 Schematic diagram of a transition sequence for estimating a first minimum initial identifier in an embodiment. DETAILED DESCRIPTION

[0028] All features disclosed in this specification, or all steps in the disclosed methods or processes, except mutually exclusive features and / or steps, can be combined in any manner.

[0029] Any feature disclosed in this specification (including any appended claims and abstract), unless otherwise stated, may be replaced by other equivalent or similar features. That is, unless otherwise stated, each feature is only an example of a series of equivalent or similar features.

[0030] Automated manufacturing systems require minimal initial resource allocation while ensuring the completion of their target tasks. Currently, the vast majority of research focuses on minimal resource allocation, with little attention paid to the operational robustness of automated manufacturing systems. Alternatively, the robustness of automated manufacturing systems is independently focused on, without integrating it with resource allocation. The present invention provides an initial resource allocation method, apparatus, and device for an automated manufacturing system, aiming to study the robust minimal initial resource allocation problem for automated manufacturing systems and simultaneously address both the minimal initial resource allocation and operational robustness issues of automated manufacturing systems.

[0031] like Figure 1 As shown, the initial resource allocation method for the automatic manufacturing system provided in the embodiment of the present application includes the following steps:

[0032] S1. Perform LPN modeling on the automated manufacturing system to obtain a labeled Petri net. If the modeled labeled Petri net contains only observable transitions, proceed to step S2; if the labeled Petri net contains unobservable transitions, proceed to step S3. In other words, the method of this embodiment is applicable to the minimum initial identity estimation problem for both labeled Petri nets containing only observable transitions and labeled Petri nets containing unobservable transitions.

[0033] The LPN modeling of the automatic manufacturing system, that is, the LPN modeling of its parts processing line, is aimed at obtaining the minimum initial resource allocation plan of the automatic manufacturing system and ensuring the robustness of the system.

[0034] S2. Estimate a first minimum initial identifier and a second minimum initial identifier in the label Petri net according to the label sequence indicating the target task / process.

[0035] The first minimum initial identifier enables at least one transition sequence corresponding to the label sequence, and the second minimum initial identifier enables all transition sequences corresponding to the label sequence.

[0036] Assume that the labeled Petri net for modeling the automatic manufacturing system is represented as (N, L, θ), where N represents the Petri net, L represents the set of letter labels, and θ is the labeling function of the transition. Given a label sequence ω, at least one transition sequence corresponding to ω can be enabled under the initial identification, and the total number of tokens in the initial identification is the least. Then this initial representation is called the first minimum initial identification.

[0037] Given the same label sequence ω, if all transition sequences corresponding to ω can be enabled in the initial identifier, and the total number of tokens in the initial identifier is the least, then the initial representation is called the second smallest initial identifier.

[0038] Both the first and second minimum initial identification schemes ensure that the automated manufacturing system meets the requirements of the label sequence ω (i.e., the target task / process) with minimal initial resource allocation, ensuring the robustness of the automated manufacturing system's operation. The difference between the two schemes lies in their ability to withstand the influence of external uncertainties. Comparing the two, the first minimum initial identification scheme is a weakly robust minimum initial resource allocation scheme, while the second minimum initial identification scheme is a strongly robust minimum initial resource allocation scheme.

[0039] As an optional implementation, the method for estimating the first minimum initial identifier in the label Petri net according to the label sequence includes:

[0040] S21. Convert the label Petri net into a label synthesis net, and use the label synthesis net to evaluate the upper limit of the minimum number of tokens of the label sequence.

[0041] For the estimation problem of the first minimum initial identifier, the label synthesis network is used to predict the minimum token upper bound of the initial identifier. The process is represented by Algorithm 1. The data structure D = (y, M) is used to capture the node information of each labeling stage, where Dy is the occurrence number vector associated with a label transition sequence observed so far, and DM is the initial identifier associated with Dy.

[0042]

[0043] S22: based on the upper limit of the minimum token number, prune the transition sequence corresponding to the label sequence, and estimate the first minimum initial identifier according to the remaining transition sequences.

[0044] The process of estimating the first minimum initial identifier based on the upper limit of the minimum token number can refer to Algorithm 2. For the data structure D2=(y,{M maim}) is used to capture the node information of the label phase (one label corresponds to one phase), where D2.y is the occurrence vector of one or more transition sequences (i.e., consistent with the label sequence observed so far), D2.{M maim} is a minimal set of initial identifiers related to D2.y, where D2.{M maim Each element of} corresponds to at least one transition sequence related to D2.y.

[0045]

[0046]

[0047] Through Algorithm 2, the minimum initial identifier corresponding to a given tag sequence, ie, the first minimum initial identifier, can be estimated under the constraint of the minimum token number upper limit MIN.

[0048] As can be seen from Algorithm 2, pruning a transition sequence is an operation of discarding some nodes in the state transition. As an optional implementation, the pruning method for the transition sequence includes:

[0049] Calculate the occurrence number vector and the corresponding initial identifier under each transition corresponding to each label in the label sequence respectively; write the initial identifiers corresponding to the same occurrence number vector into the same identifier set in sequence.

[0050] When writing a new initial identifier for the same identifier set (i.e., in the same tag phase), each initial identifier in the identifier set is traversed. If the currently traversed initial identifier exceeds the initial identifier to be written, that is, the currently traversed transition sequence has more tokens than the newly traversed transition sequence, the currently traversed initial identifier is discarded and the initial identifier to be written is written. Transition sequences with tokens exceeding the minimum token limit are directly discarded and not considered.

[0051] As an optional implementation, the method for estimating the second minimum initial identifier in the label Petri net according to the label sequence includes:

[0052] S23. Estimate the occurrence number vector and the corresponding minimum initial identifier of each transition sequence corresponding to the label sequence using the label Petri net.

[0053] S24. Take the union of all estimated minimum initial identifiers to obtain the second minimum initial identifier.

[0054] The second minimum initial identifier enables all transition sequences corresponding to the label sequence. Therefore, in an embodiment of the present application, an enumeration operation is performed on each transition sequence corresponding to the label sequence based on an iterative algorithm, and the corresponding identifiers are taken as a union to estimate the second minimum initial identifier.

[0055] With S = {t 11 t 12 ...t 1k ,t 21 t 22 ...t 2k ,...,t d1 t d2 ...t dk} represents the given label sequence ω=l1l2...l k The set of all corresponding transition sequences, where d = |S| represents the number of elements in the set, that is, the number of transition sequences.

[0056] According to Algorithm 3, the second minimum initial identifier can be estimated.

[0057]

[0058] S3. Estimate a first minimum initial identifier in the label Petri net according to the label sequence.

[0059] For a labeled Petri net containing unobservable transitions, in an embodiment of the present application, a minimum initial identifier that enables at least one transition sequence corresponding to the label sequence is estimated.

[0060] For labeled Petri nets containing unobservable transitions, the concepts of minimum interpretation and minimum interpretation place are involved when estimating the first minimum initial identification.

[0061] In a labeled Petri net (N,L∪{ε},θ), N=(P,T,F,W), for t∈T o (observable transition set), marked M∈R(N,M0), where M0 is an initial mark, the interpretation set of transition t under mark M is:

[0062] Among them, T u Identifies the set of unobservable transitions.

[0063] That is, the interpretation set of transition t under the label M is the set of unobservable transitions that can indirectly enable transition t under the label M.

[0064] The occurrence vector set under E(M,t) is expressed as:

[0065] Y(M,t)={π(σ)∈N m |σ∈E(M,t)},

[0066] Where N m Represents an m-dimensional integer space.

[0067] The minimum interpretation set of transition t under the symbol M is expressed as:

[0068]

[0069] In E min The occurrence vector set under (M, t) is expressed as:

[0070] Y min (M,t)={π(σ)∈N m |σ∈E min (M,t)}.

[0071] For t∈T o , M∈R(N,M0), R(N,M0) represents the reachable graph of Petri net N in the initial state M0. If M[t>, then we can get ε∈E(M,t), where ε represents a transition sequence of length 0, and we can also get therefore like Let M[σ>M′[t>, then:

[0072]

[0073] The method for obtaining the occurrence number vector under the minimum explanation can refer to Algorithm 4.

[0074]

[0075]

[0076] like Then the set of interpretation libraries of transition t under the label M is expressed as:

[0077] Represents a collection of interpretation libraries.

[0078] The minimum set of interpretation libraries for transition t under the symbol M is expressed as:

[0079]

[0080] The potential minimum interpretation set of transition t under the label M is expressed as:

[0081]

[0082] In general, it is difficult to find the minimum set of explanatory places for the LPN model due to the complex combination of related positions. To improve the analyzability of the model, this application, in some optional implementations, transforms the label Petri net. When calculating the minimum explanatory place for observable transitions, the Petri net is constrained as follows:

[0083] For any unobservable transition, the number of input places is 1. That is, for t∈T u (unobservable transition set), |·t|=1.

[0084] For any observable transition, if its input place is more than 1, then the front set of its input place and the set of unobservable transitions do not intersect. That is, for t∈T o , if |·t|≥2, then

[0085] The method for obtaining the relevant library place set in the labeled Petri net for transition t is shown in Algorithm 5.

[0086]

[0087] Get the observable change t at the current marker M cur The method for the minimum interpretation library under can be found in Algorithm 6.

[0088]

[0089]

[0090] The previous article described how to obtain the minimum interpretation and minimum interpretation place for observable transitions. Based on the minimum interpretation and minimum interpretation place, the infinite solution space of the excitation sequence of unobservable transitions is transformed into a finite solution space. At the same time, the Petri net structure is required to be an unobservable subnet with no transition connections. Based on the constraints on labeled Petri nets described above, the first minimum initial identification of the labeled Petri net containing unobservable transitions is estimated.

[0091] In some optional embodiments, for each observable transition corresponding to the label sequence in the label Petri net, the minimum interpretation and the minimum interpretation library under the current initial identification are calculated; the initial identification is updated according to the requirement that only the minimum interpretation or the potential minimum interpretation obtained from the minimum interpretation library is allowed to be emitted before each observable transition corresponding to the label sequence is emitted; this cycle is repeated to obtain the first minimum initial identification.

[0092] In some optional implementations, updating the initial identifier includes:

[0093] S31. For the observable changes of the label sequence in each stage in the label Petri net, calculate the minimum interpretation and the minimum interpretation library under the initial identification of the stage.

[0094] S32. According to the minimum explanation and minimum explanation library of each observable transition, the next observable transition after each observable transition is emitted is calculated to obtain an observable transition sequence.

[0095] Specifically, when the minimum explanation of the observable change exists, the next observable change is calculated based on the occurrence number vector under the minimum explanation; when the minimum explanation library of the observable change exists, the next observable change is calculated based on the minimum explanation library; when neither the minimum explanation of the observable change nor the minimum explanation library exists, the next observable change corresponding to all label sequences is calculated.

[0096] S33: Prune the observable transition sequences, and estimate the first minimum initial identifier based on the remaining observable transition sequences.

[0097] The pruning method for observable transition sequences is the same as the previous pruning method, namely:

[0098] The occurrence number vector and the corresponding initial identification under each observable transition corresponding to each label in the label sequence are calculated respectively; and the initial identifications corresponding to the same occurrence number vector are sequentially written into the same identification set.

[0099] For the same identifier set, when writing a new initial identifier, traverse each initial identifier in the identifier set. If the currently traversed initial identifier exceeds the initial identifier to be written, discard the currently traversed initial identifier and write the initial identifier to be written.

[0100] The method for implementing the above-mentioned first minimum initial identification estimation can be completed with reference to Algorithm 7.

[0101]

[0102]

[0103] S4. Allocate the resource indicated by the obtained minimum initial identifier to the target task / process.

[0104] In step S4, the first minimum initial identifier and the second minimum initial identifier obtained in step S2, or the first minimum initial identifier obtained in step S3, are allocated initial resources according to the values of the identifiers.

[0105] For example, if Figure 2 The following is a network model of the labeled Petri net obtained by LPN modeling of an automatic manufacturing system. This is a labeled Petri net that only contains observable transitions. The labeled synthesis net converted from this labeled Petri net is shown in Figure 3 For the definition of each node in the label Petri net, please refer to Table 1.

[0106] Table 1 Node definition table

[0107] Node (Place / Transition) Physical definition <![CDATA[p1]]> Parts 1 Library <![CDATA[p2]]> Parts 2 Library <![CDATA[p3]]> Staging Library <![CDATA[p4]]> Residue 1 <![CDATA[p5]]> Residue 2 <![CDATA[p6]]> Finished product warehouse <![CDATA[t1]]> Processing of parts 1 and 2 <![CDATA[t2]]> Recasting of the remaining material 1 <![CDATA[t3]]> Recasting of the remaining material <![CDATA[t4]]> Quality inspection and packaging of products <![CDATA[t5]]> finishing

[0108] When a tag sequence ω=abba is given, the first minimum initial identifier and the second minimum initial identifier are estimated respectively through the method of step S2.

[0109] For the estimate of the second minimum initial identifier, such as Figure 4 The following table shows all the transition sequences corresponding to the label sequence ω = abba. The values calculated for each node are shown in Table 2.

[0110] Table 2 Node information table

[0111]

[0112]

[0113] from Figure 4 It can be seen that from the node (y 15 ,{M 15}) to node (y 30 ,{M 30}) are the occurrence vectors of the transition sequence corresponding to the label sequence and the minimum initial identifier. By taking the union of the minimum initial identifiers of these nodes, the second minimum initial identifier M is obtained. SRMuIM =[212220] T .

[0114] Similarly, through the method of step S2, the upper limit of the minimum number of tokens of the tag sequence ω=abba is calculated to be MIN=2.5, and the transition sequence is pruned using MIN. Figure 5 As shown, based on MIN, delete Figure 4 Nodes (y7,{M7}), (y 10 ,{M 10})、(y 11 ,{M 11})、(y 12 ,{M 12})、(y 13 ,{M 13}) and (y 14 ,{M 14}), complete the pruning of the transition sequence, which can save retrieval time and improve search efficiency. Finally, the first minimum initial identifier is obtained as node (y 18 ,{M 18}) and node (y 20 ,{M 20}) The initial identifier corresponding to M WRMuIM =[110000] T For nodes with the same occurrence vector, the corresponding initial identifiers will be compared and judged to form a set, which is expressed as D2=(y,{M maim}).

[0115] like Figure 6 The following table shows the network model of the Labeled Petri Net obtained by LPN modeling of another automatic manufacturing system. The definition of each node is shown in Table 3.

[0116] Table 3 Node definition table

[0117] Repository / Change Practical significance <![CDATA[p1]]> The repository where artifacts are stored <![CDATA[p2]]> The workpiece is processed by machine No. 1 <![CDATA[p3]]> The workpiece is processed by machine No. 2 <![CDATA[p4]]> The workpiece is processed by machine No. 3 <![CDATA[p5]]> The workpiece is processed by machine No. 4 <![CDATA[p6]]> Finished product warehouse <![CDATA[t1]]> The robot loads the workpiece onto machine No. 2 <![CDATA[t2]]> The robot loads the workpiece into machine No. 1 <![CDATA[t3]]> The robot loads the workpiece onto machine No. 4 <![CDATA[t4]]> The robot loads the workpiece onto machine No. 3 <![CDATA[t5]]> The finished products are loaded into the finished product warehouse by robots <![CDATA[t6]]> The robot moves part of the workpiece from machine No. 2 to machine No. 1 <![CDATA[t7]]> The finished products are loaded into the finished product warehouse by robots

[0118] Given a label sequence ω=aa, the corresponding transition sequences in the label Petri net are as follows Figure 7 shown. Figure 7 The node information of each node in is shown in Table 4.

[0119] Table 4 Node information table

[0120] node Node Information <![CDATA[(y1,{M1})]]> <![CDATA[([0000000] T ,{[000000] T })]]> <![CDATA[(y2,{M2})]]> <![CDATA[([1000000] T ,{[100000] T })]]> <![CDATA[(y3,{M3})]]> <![CDATA[([0001000] T ,{[010000] T })]]> <![CDATA[(y4,{M4})]]> <![CDATA[([0001010] T ,{[001000] T })]]> <![CDATA[(y5,{M5})]]> <![CDATA[([2000000] T ,{[200000] T })]]> <![CDATA[(y6,{M6})]]> <![CDATA[([1001010] T ,{[100000] T })]]> <![CDATA[(y7,{M7})]]> <![CDATA[([1001000] T ,{[110000] T })]]> <![CDATA[(y8,{M8})]]> <![CDATA[([0002000] T ,{[020000] T })]]> <![CDATA[(y9,{M9})]]> <![CDATA[([0002010] T ,{[011000] T })]]> <![CDATA[(y 10 ,{M 10 })]]> <![CDATA[([1001010] T ,{[101000] T })]]> <![CDATA[(y 11 ,{M 11 })]]> <![CDATA[([0002010] T ,{[011000] T })]]> <![CDATA[(y 12 ,{M 12 })]]> <![CDATA[([0002020] T ,{[002000] T })]]>

[0121] By the method of step S3 above, at the current mark M i +[N]·y i (i=1,2,3,4) to find the minimum explanation E for the observable change t min (M i +[N]·y i ,t) and the potential minimum explanation E pmin (M i +[N]·y i ,t), the calculation results are shown in Table 5.

[0122] Table 5 Minimum explanation and potential minimum explanation

[0123]

[0124]

[0125] Since the label sequence ω=aa contains only two labels and no more labels are observed after aa, there is no need to calculate the minimum explanation set and potential minimum explanation set of the subsequent nodes.

[0126] Depend on Figure 7 It can be seen that the different transition sequences corresponding to the label sequence ω=aa are t1t1, t1t6t4, t4t1, t4t4, t4t6t4, t6t4t1, t6t4t4 and t6t4t6t4. Using the method of step S3 to operate on different transition sequences, the first minimum initial identification set is finally obtained: Z WRMuIM (ω)={[1 0 0 0 0 0] T}.

[0127] At the end of the above two specific implementation cases, according to the obtained M SRMuIM和 M WRMuIM , or the obtained Z WRMuIM , perform initial resource allocation.

[0128] According to the concept of the present application, an initial resource allocation device for an automatic manufacturing system is also proposed in an embodiment of the present application. The device includes:

[0129] The first module is used to perform LPN modeling on the automatic manufacturing system to obtain a label Petri net; if the label Petri net contains only observable changes, the second module is activated; if the label Petri net contains unobservable changes, the third module is activated.

[0130] The second module is configured to estimate a first minimum initial identifier and a second minimum initial identifier in a label Petri net based on a label sequence indicating a target task / process. The first minimum initial identifier enables at least one transition sequence corresponding to the label sequence, and the second minimum initial identifier enables all transition sequences corresponding to the label sequence.

[0131] The third module is used to estimate a first minimum initial identifier in the label Petri net according to the label sequence.

[0132] The fourth module allocates the resources indicated by the obtained minimum initial identifier to the target task / process.

[0133] The data configured for each module of the above-mentioned device can refer to the features designed for each step in the above-mentioned method embodiment.

[0134] In addition, an embodiment of the present application also provides an initial resource allocation device for an automatic manufacturing system, which includes a processor and a storage medium, wherein the storage medium stores computer instructions. When the processor runs the computer instructions, it can execute the initial resource allocation method for the automatic manufacturing system of the above embodiment.

[0135] The present invention is not limited to the aforementioned specific embodiments, but extends to any new features or any new combination disclosed in this specification, as well as any new method or process steps or any new combination disclosed.

Claims

1. An initial resource allocation method for an automatic manufacturing system, characterized in that: include: S1. Perform LPN modeling on the automatic manufacturing system to obtain a labeled Petri net; if the labeled Petri net contains only observable transitions, execute step S2; if the labeled Petri net contains unobservable transitions, execute step S3; S2. Estimate a first minimum initial identifier and a second minimum initial identifier in a label Petri net based on a label sequence indicating a target task / process; the first minimum initial identifier enables at least one transition sequence corresponding to the label sequence, and the second minimum initial identifier enables all transition sequences corresponding to the label sequence; S3. Estimating a first minimum initial identifier in a label Petri net according to the label sequence; S4. Allocate the resource indicated by the obtained minimum initial identifier to the target task / process.

2. The method for initial resource allocation of an automatic manufacturing system according to claim 1, wherein: Based on the label sequence indicating the target task / process, the first minimum initial identifier is estimated in the label Petri net, including: S21, converting the label Petri net into a label synthesis net, and using the label synthesis net to evaluate the upper limit of the minimum token number of the label sequence; S22: Prune the transition sequence corresponding to the label sequence based on the upper limit of the minimum token number, and estimate the first minimum initial identifier according to the remaining transition sequences.

3. The method for initial resource allocation of an automatic manufacturing system according to claim 1, wherein: According to the label sequence indicating the target task / process, the second minimum initial identification is estimated in the label Petri net, including: S23, estimating the occurrence number vector and the corresponding minimum initial identifier of each transition sequence corresponding to the label sequence using the label Petri net; S24. Take the union of all estimated minimum initial identifiers to obtain the second minimum initial identifier.

4. The method for initial resource allocation of an automatic manufacturing system according to claim 1, wherein: In step S3, estimating a first minimum initial identifier in a label Petri net according to the label sequence includes: For each observable transition corresponding to the label sequence in the label Petri net, the minimum interpretation and the minimum interpretation library under the current initial identifier are calculated; the initial identifier is updated according to the requirement that only the minimum interpretation or the potential minimum interpretation obtained from the minimum interpretation library is allowed to be emitted before each observable transition corresponding to the label sequence is emitted; this cycle is repeated to obtain the first minimum initial identifier.

5. The method for initial resource allocation of an automatic manufacturing system according to claim 4, wherein: The method for updating the initial identifier includes: S31, for the observable changes corresponding to the label sequence at each stage in the label Petri net, calculate the minimum interpretation and minimum interpretation place under the initial identification of the stage; S32. Calculate the next observable transition after each observable transition is emitted based on the minimum interpretation and minimum interpretation library of each observable transition to obtain an observable transition sequence; S33: Prune the observable transition sequence, and estimate the first minimum initial identifier based on the remaining observable transition sequences.

6. The method for initial resource allocation of an automatic manufacturing system according to claim 5, wherein: According to the minimum explanation of each observable transition and the existence of the minimum explanation library, the next observable transition after each observable transition is emitted is calculated, including: When the minimum explanation of an observable change exists, the next observable change is calculated based on the occurrence number vector under the minimum explanation; When the minimum explanation base of the observable change exists, the next observable change is calculated according to the minimum explanation base; When neither the minimum explanation nor the minimum explanation library of the observable transition exists, all next observable transitions corresponding to the tag sequence are calculated.

7. The method for initial resource allocation of an automatic manufacturing system according to claim 5 or 6, characterized in that: When calculating the minimum explanation place of observable transitions, the following constraints are imposed on the Petri net: For any unobservable transition, the input place is 1; For any observable transition, if its input places are more than one, then the front set of its input places does not intersect with the set of unobservable transitions.

8. The method for initial resource allocation of an automatic manufacturing system according to claim 2 or 5, characterized in that: Methods for pruning transition sequences include: Calculating the occurrence number vector and the corresponding initial identifier under each transition corresponding to each label in the label sequence respectively; writing each initial identifier corresponding to the same occurrence number vector into the same identifier set in sequence; For the same identifier set, when writing a new initial identifier, traverse each initial identifier in the identifier set. If the currently traversed initial identifier exceeds the initial identifier to be written, discard the currently traversed initial identifier and write the initial identifier to be written.

9. An initial resource allocation device for an automatic manufacturing system, characterized in that: include: The first module is used to perform LPN modeling on the automatic manufacturing system to obtain a labeled Petri net; if the labeled Petri net contains only observable transitions, the second module is activated; if the labeled Petri net contains unobservable transitions, the third module is activated; The second module is configured to estimate a first minimum initial identifier and a second minimum initial identifier in a label Petri net based on a label sequence indicating a target task / process; the first minimum initial identifier enables at least one transition sequence corresponding to the label sequence, and the second minimum initial identifier enables all transition sequences corresponding to the label sequence; A third module is configured to estimate a first minimum initial identifier in a label Petri net according to the label sequence; The fourth module allocates the resources indicated by the obtained minimum initial identifier to the target task / process.

10. An initial resource allocation device for an automatic manufacturing system, characterized in that: The method comprises a processor and a storage medium, wherein the storage medium stores computer instructions. When the processor runs the computer instructions, the method for initial resource allocation of an automatic manufacturing system according to any one of claims 1 to 8 can be executed.

Images