Intelligent manufacturing system language processing method based on collaborative awareness

Abstract

The invention provides an intelligent manufacturing system language processing method based on collaborative awareness, and the method comprises the steps: employing a data management convergence factor based on a language utility function under a static condition; under a dynamic condition, a mode of combining a data management convergence factor based on a language utility function and a particle swarm optimization feedback parameter is adopted to carry out language processing and generation task processing and resource redistribution, so that intelligent manufacturing system language processing based on collaborative awareness is realized. The key scientific problems of the intelligent manufacturing system in the aspects of intelligent modeling, accurate semantic analysis, Chinese language perception, intelligent language generation and the like are solved, and theoretical support and application guidance are provided for improving the man-machine interaction of the intelligent manufacturing system.

Description

Technical Field

[0001] The present invention belongs to the field of language processing, and in particular relates to a language processing method for an intelligent manufacturing system based on collaborative perception. Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] The core of the language processing technology of the intelligent manufacturing system is a multi-hop wireless self-organizing network system composed of a large number of language-aware nodes deployed in the working area. Each node collaboratively perceives, collects and processes the language information of the perceived objects in the coverage area of ​​the intelligent manufacturing system.

[0004] At present, how to coordinately perceive, collect and process various language application object information in the network coverage area, conduct accurate semantic analysis, and publish it to the intelligent manufacturing system functional modules that need this information has become a primary problem that needs to be solved. The relevant theories, solution frameworks and technologies of language processing and generation are gradually becoming research hotspots. If we can use the language adaptive collaborative perception theory to explore its key technologies and applications of human-computer interaction in intelligent manufacturing systems, it will provide necessary support for improving the intelligence and interactivity of intelligent manufacturing systems.

[0005] However, due to the complexity of the intelligent manufacturing system itself, there is currently no mature technology for the application of its language processing and generation technology. Summary of the invention

[0006] In order to solve the above problems, the present invention proposes a language processing method for an intelligent manufacturing system based on collaborative perception. The present invention solves the key scientific problems of the intelligent manufacturing system in intelligent modeling, precise semantic analysis, Chinese language perception and intelligent language generation, and provides theoretical support and application guidance for improving the human-computer interactivity of the intelligent manufacturing system.

[0007] According to some embodiments, the present invention adopts the following technical solutions:

[0008] A language processing method for an intelligent manufacturing system based on collaborative perception comprises the following steps:

[0009] Initialize language task data, including total task completion time, energy consumption and network load status;

[0010] According to the initialized data, a cost function is constructed and optimized using the particle swarm optimization algorithm. Based on the mapping relationship between particles and the nonlinear characteristics of the position and velocity of each particle, the data set of the system language model is obtained.

[0011] The collaborative group is established based on the nonlinear characteristics of the language data set, and the numerical characteristics of the collaborative group are set by using the language cohesion and semantic relevance to ensure the realization of the node's ability to perceive language processing requirements individually;

[0012] Based on the principle that vectors are orthogonal to each other and there is no direct mapping, the idea of ​​joint optimization of language attributes and collaborative group feature vectors is used to make joint judgments and estimates of data of related functional nodes within the group.

[0013] Establish a system node collaborative coverage model and set constraints for the node collaborative coverage model;

[0014] Based on the collaborative coverage model, a dynamic balancing scheme is used to improve the connection strategy between system language generation tasks;

[0015] Configure the characteristics of system routing self-adjustment and self-configuration in the case of multi-source language fusion, and perform intra-cluster coverage control based on the characteristics;

[0016] From the perspective of multi-source similar information fusion, the language generation requirements with high correlation and strong spatiotemporal correlation are integrated;

[0017] Under static conditions, a data management convergence factor based on language utility function is adopted; under dynamic conditions, a combination of a data management convergence factor based on language utility function and a particle swarm optimization feedback parameter is adopted to carry out language processing, generate task processing and resource redistribution, and realize language processing of intelligent manufacturing system based on collaborative perception.

[0018] As an optional implementation, the process of constructing the cost function includes:

[0019] The cost function is constructed with the minimum system energy consumption as the objective function and the maintenance of network semantic node connectivity as the constraint condition.

[0020] As an optional implementation method, the process of establishing a system node cooperative coverage model and setting constraints of the node cooperative coverage model includes:

[0021] Based on the requirements of system network logic mapping coverage and energy consumption minimization, on plane A containing n network semantic nodes n The spatial constraint mechanism is used, and the logical mapping radius of each network semantic node is d; the network nodes are on plane A n The uniform distribution is followed, and the total semantic interference of the network is reduced by increasing the plane.

[0022] As an optional implementation, based on the collaborative coverage model, the process of improving the connection strategy between system language generation tasks using a dynamic balancing solution includes:

[0023] Let l i,l j ∈R 2 Represents the network semantic node v i ,v j The position of v i ≠v j ; Define network semantic node v i and v j The direct connectivity is d; when there is a non-empty set in the network node group P, set the network semantic node v i and v j There are multi-hop connections between i and v j The information between can be effectively addressed and transmitted through the node group P; given the number of nodes is n, let l scr ,l 1 ,l 2 ,...,l n-1 ∈R 2 Represent the source node and node v respectively 1 ,v 2 , …, v -1 , and let V be the set of all network semantic nodes; set N(l i ) is l i The maximum number of nodes in the logical mapping range with d as the center and d as the mapping radius, and N(l i )={v z :v z ∈Vand‖l i -l z ‖≤d}; all network semantic nodes have the same logical mapping coverage radius and mapping correlation. The adjustability of system nodes is used to improve the coverage quality of sensitive area monitoring and realize inter-node collaboration and adjustment under energy-constrained conditions.

[0024] As an optional implementation method, the process of configuring the self-adjustment and self-configuration characteristics of system routing in the case of multi-source language fusion and performing the logic mapping coverage control within the cluster includes:

[0025] When merging multiple source languages, without considering the semantic boundary effect, the third-party network node v k Able to communicate with network nodes v j Achieve direct interconnection but cannot communicate with network nodes v i The upper bound for achieving direct interconnection is A(v j )-[A(v i )∩A(v j )], thereby realizing the characteristics of self-adjustment and self-configuration of system routing. Set the existence positions to be l i ,l j ∈R2 Network nodes v i and v j , and ||l i -l j ||=d; ignoring the additional space of the last network node, the lower bound of the spatial position of the network node is Let T src is the total number of transmissions required when using the multi-hop method under error-free conditions, assuming that n network nodes are randomly located in area A according to uniform distribution n Within, the logical mapping coverage radius is d. When a network node transmits T = k times to realize a single bit broadcast to other (n-1) network nodes, its probability upper bound is G(k,n)(d 2 / A n ) n-1 ,in and This enables logical mapping coverage control within the cluster.

[0026] As an optional implementation method, from the perspective of multi-source similar information fusion, the process of fusing language generation requirements with high correlation and strong spatiotemporal correlation includes:

[0027] Based on the relationship between semantic nodes and the semantic stability of the system, the probability of RTS initiated by the semantic node of the intelligent manufacturing system network for flow f is defined as P nf , define the probability of successful data transmission for flow f as P sf , initialize the energy consumption and load status of semantic nodes;

[0028] According to the initialized parameters, a cost function is constructed in the system perception strategy process based on precise semantic analysis, and the particle swarm optimization algorithm is used for optimization to obtain the intermediate state analysis model of the semantic node;

[0029] According to the obtained semantic node intermediate state analysis model, a collaborative group of source semantic node set and target semantic node set is established to model the data collaborative perception evolution and obtain a system perception strategy based on precise semantic analysis.

[0030] According to the obtained precise semantic analysis system perception strategy, based on the language generation mechanism, the channel status of the source semantic node before initiating RTS is dynamically controlled, and the semantic relevance and spatiotemporal correlation of the destination semantic node are dynamically adjusted to ensure the fusion performance of multi-source language generation requirements;

[0031] From the perspective of multi-source similar information fusion, the language generation requirements with correlation greater than the set value and spatiotemporal correlation intensity exceeding the preset value are integrated.

[0032] As a further step, the particle swarm optimization algorithm is used to optimize and obtain the intermediate state analysis model of the semantic node, including:

[0033] Define the probability that the destination semantic node becomes a hidden node due to exceeding the listening range of the source semantic node as Y t , in the steady state Y t With P nf and P sf The relationship is P sf With P nf The ratio of represents the source semantic node's listening efficiency and data correctness probability for stream f, and the basic framework of the semantic node intermediate state analysis model is characterized by the source semantic node listening effectiveness;

[0034] Based on the particle swarm optimization algorithm, the maximum number of RTS initiated by the network before the data packet is discarded is set to u r , construct the state transition process of the intermediate state analysis model of the semantic node.

[0035] As a further step, according to the obtained semantic node intermediate state analysis model, a collaborative group of a source semantic node set and a target semantic node set is established, and the process of data collaborative perception evolution modeling includes:

[0036] Set Y t Indicates the probability of a hidden node appearing each time a semantic node initiates RTS, 1-Y t It indicates the probability of a data packet achieving normal addressing each time an RTS is initiated; based on the semantic node intermediate state analysis model, the mapping relationship between the source semantic node set and the destination semantic node set is abstracted, and the coordination group is constructed using the nonlinear mapping between the sets;

[0037] By utilizing the vector characteristics and spatiotemporal correlation of the collaborative group, the number of times the semantic network initiates RTS and fails is controlled within a reasonable range. Based on the performance requirements of multi-information fusion perception, a system perception strategy based on precise semantic analysis is obtained.

[0038] As an optional implementation method, under static conditions, a data management convergence factor based on a language utility function is adopted; under dynamic conditions, a method combining a data management convergence factor based on a language utility function and a particle swarm optimization feedback parameter is adopted to perform language processing and generation task processing and resource redistribution, and the process includes:

[0039] Taking advantage of the fact that mathematical space is a locally compact space, the perception strategy model of the precise semantic analysis system based on static conditions and dynamic conditions is optimized respectively, and a solution of a multi-objective optimization genetic algorithm is designed according to the characteristics of the locally compact space.

[0040] The adjustability of semantic nodes is used to improve the coverage quality of language-sensitive area monitoring, and the characteristics of language utility functions are used to achieve collaboration and adjustment between semantic nodes under energy-constrained conditions;

[0041] The probability lower bound of the network connectivity is set through the Poisson Boolean model, and the maximum value of semantic elements transmitted simultaneously by the system and the language processing limit are set;

[0042] Set the system comprehensive packet loss rate, and perform language processing and generation task processing and resource redistribution based on the maximum value of semantic elements transmitted simultaneously by the system and the language processing limit value, based on the data management convergence factor of the language utility function and the particle swarm optimization feedback parameter, to meet the language collaborative perception needs of all nodes in the system;

[0043] Based on the initiation characteristics of multi-source languages ​​and the sensitivity of data collaborative perception mechanisms, the accuracy of language collaborative perception decisions is described;

[0044] Based on the accurate description of language collaborative perception decision-making, language processing, task generation processing and resource reallocation processing are realized to complete the language processing of the intelligent manufacturing system based on collaborative perception.

[0045] As a further step, the process of setting the probability lower bound of the connectivity of the network through the Poisson Boolean model and setting the maximum value of the semantic elements transmitted simultaneously by the system and the language processing limit value includes:

[0046] Set the probability of connecting semantic nodes in the interval 0≤x≤D to P oD (x) = P o (x)+P o (Dx)-P o (x)P o (Dx), where The average connection probability of a semantic node occupying the language utility function is P oD The lower bound of depends on the worst state of the node position, that is, P oD+ =inf 0≤x≤D {P oD (x)}=2P o (D / 2)-P o 2 (D / 2); because P o The lower bound of (x) is So when 2r≤D≤4r, P o (D / 2) = 1-e -λr -λDe -λr / 2 and P oD The lower bound of P oD+ =1-(λD / 2+1) 2 e -2λr ; Based on the above characteristics, the data management convergence factor of the language utility function is used to improve the convergence stability of language processing;

[0047] Set the average number of semantic nodes sent in a stable state n y =b l n t , where b l is the network congestion coefficient. To ensure the normal initiation of RTS, the minimum space requirement is Among them, M c is the carrier sense capacity, which indicates the maximum number of RTSs that can be initiated without packet collision. Combined with the particle swarm optimization feedback parameter, we have nodes can successfully transmit data. RTS initiated simultaneously, of which M b is the data transmission capacity. Based on the above characteristics, the maximum value of the semantic elements that the system can transmit simultaneously and the language processing limit of the system are obtained.

[0048] Compared with the prior art, the present invention has the following beneficial effects:

[0049] The present invention deeply combines the adaptive collaborative perception theory with the language processing technology of the intelligent manufacturing system. The multi-source language processing technology based on the adaptive collaborative perception theory and high accuracy proposed by the present invention can effectively promote the intelligence and interactivity of the intelligent manufacturing system.

[0050] The present invention proposes an algorithm for calculating the fault tolerance and scalability of language aggregation functions in the collaborative perception subsystem of the multi-source language adaptive collaborative perception solution of the intelligent manufacturing system, determines the node energy consumption parameters and load indicators, and solves the problem of applying language processing technology to language perception data query processing.

[0051] The present invention designs a scheduling algorithm that supports high reliability strategy, elastic computing strategy, dynamic resource allocation strategy and disaster recovery backup strategy. It achieves the goal of detecting the language generation capability of nodes in a certain direction by collaboratively controlling language information in a hop-by-hop routing manner, and proposes a strategy for updating the intelligent manufacturing system model and optimizing language generation.

[0052] The present invention uses language feedback mechanism and particle swarm optimization algorithm to establish a system-oriented multi-source Chinese language fusion model. A new Chinese language feedback prediction algorithm is constructed to perform numerical prediction by predicting grammar and semantics, thus achieving the purpose of further improving the interactivity of the intelligent manufacturing system.

[0053] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0055] Figure 1 is a general design diagram of language processing of an intelligent manufacturing system of an embodiment;

[0056] Figure 2 is a schematic diagram of a language processing process of an intelligent manufacturing system based on collaborative perception according to an embodiment;

[0057] Figure 3 A relationship diagram between a current node and neighboring nodes in an adaptive collaborative perception semantic analysis model of an embodiment;

[0058] Figure 4 The present invention is a schematic diagram of the state transition process of a system perception strategy based on precise semantic analysis in an embodiment. DETAILED DESCRIPTION

[0059] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0060] It should be noted that the following detailed descriptions are all illustrative and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0061] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0062] In the absence of conflict, the embodiments in this application and the features in the embodiments may be combined with each other.

[0063] Embodiment 1

[0064] As described in the background technology, for the key technologies and applications of human-computer interaction in intelligent manufacturing systems, in order to solve the problems of low system adaptive cognitive ability, unbalanced load, inaccurate semantic analysis, weak predictability of multi-source Chinese language, and insufficient dynamics of language generation balancing algorithm, this embodiment provides a language processing method for intelligent manufacturing systems based on collaborative perception to achieve the intelligence and interactivity of intelligent manufacturing systems.

[0065] A language processing method for an intelligent manufacturing system based on collaborative perception comprises the following steps:

[0066] Step S1: Initialize the total system task completion time, energy consumption and network load status;

[0067] Step S2: construct a cost function based on the initialized parameter data, optimize it using the particle swarm optimization algorithm, and obtain the data set of the system language model based on the mapping relationship between particles and the nonlinear characteristics of the position and velocity of each particle;

[0068] based on Figure 1 The data collaborative perception subsystem takes the minimum system energy consumption as the objective function and the maintenance of network semantic node continuity as the constraint condition, and constructs the same cost function in its various modules.

[0069] Step S3: Establish a collaborative group based on the nonlinear characteristics of the language data set inherited from the particle swarm, and use the stickiness of the language and the relevance of the semantics to set the numerical characteristics of the collaborative group to ensure the realization of the ability of the node to perceive the language processing requirements independently. Based on the principle that vectors are orthogonal to each other and there is no direct mapping, the idea of ​​jointly optimizing the language's own attributes and the collaborative group's feature vectors is used to make joint judgments and estimates of the data of related functional nodes within the group.

[0070] The above nonlinear characteristics refer to the nonlinearity of the mapping relationship between nodes reflected by the particle swarm algorithm.

[0071] The numerical properties of collaborative groups are set using the cohesiveness of the language and the relevance of the semantics.

[0072] Step S4: Establish a system semantic node collaborative coverage model and set constraints for the node collaborative coverage model;

[0073] Specifically, based on the requirements of system network logic mapping coverage and energy consumption minimization, on plane A containing n network semantic nodes n The spatial constraint mechanism is used, and the logical mapping radius of each network semantic node is d; the network nodes are on plane A n The uniform distribution is followed, and the total semantic interference of the network is reduced by increasing the plane.

[0074] Step S5: Based on the collaborative coverage model, a dynamic balancing solution is used to improve the connection strategy between system language generation tasks.

[0075] In this embodiment, let l i ,l j ∈R 2 Represents the network semantic node v i ,v j The position of v i ≠v j ; Define network semantic node v i and v j The direct connectivity is d; when there is a non-empty set in the network node group P, set the network semantic node v i and v j There are multi-hop connections between i and v j The information between can be effectively addressed and transmitted through the node group P; given the number of nodes is n, let l scr ,l 1 ,l 2 ,...,l n-1 ∈R 2 Represent the source node and node v respectively 1 ,v 2 , …, v -1 , and let V be the set of all network semantic nodes; set N(l i ) is l i The maximum number of nodes in the logical mapping range with d as the center and d as the mapping radius, and N(l i )={v z :v z ∈Vand||l i -l z ||≤d}; all network semantic nodes have the same logical mapping coverage radius and mapping correlation, and the adjustability of system nodes is used to improve the coverage quality of sensitive area monitoring, and to achieve collaboration and adjustment among nodes under energy-constrained conditions.

[0076] Step S6: configuring the characteristics of system routing self-adjustment and self-configuration in the case of multi-source language fusion, and performing intra-cluster coverage control based on the characteristics;

[0077] Based on Figure 2 The framework shown in the figure shows that when multiple source languages ​​are fused, without considering the semantic boundary effect, the third-party network node v k Able to communicate with network nodes v j Achieve direct interconnection but cannot communicate with network nodes v i The upper bound for achieving direct interconnection is A(v j )-[A(vi )∩A(v j )], thereby realizing the characteristics of self-adjustment and self-configuration of system routing. Set the existence positions to be l i ,l j ∈R 2 Network nodes v i and v j , and ||l i -l j ||=d; ignoring the additional space of the last network node, the lower bound of the spatial position of the network node is Let T src is the total number of transmissions required when using the multi-hop method under error-free conditions, assuming that n network nodes are randomly located in area A according to uniform distribution n Within, the logical mapping coverage radius is d. When a network node transmits T = k times to realize a single bit broadcast to other (n-1) network nodes, its probability upper bound is G(k,n)(d 2 / A n ) n-1 ,in and This enables logical mapping coverage control within the cluster.

[0078] Step S7: from the perspective of multi-source similar information fusion, the language generation requirements with high correlation and strong spatiotemporal correlation are integrated;

[0079] Specifically, the step includes:

[0080] Step S71: Based on Figure 3 The relationship between the semantic nodes shown in the figure is based on the semantic stability of the system. The probability of the semantic node of the intelligent manufacturing system network initiating RTS for flow f is defined as P nf , define the probability of successful data transmission for flow f as P sf , initialize the energy consumption and load status of semantic nodes;

[0081] Step S72: constructing a cost function in the system perception strategy process based on precise semantic analysis according to the initialized parameters, optimizing it using a particle swarm optimization algorithm, and obtaining an intermediate state analysis model of the semantic node;

[0082] Step S72 further comprises:

[0083] Step S7201: define the probability that the destination semantic node becomes a hidden node due to exceeding the listening range of the source semantic node as Y t , in the steady state Y t With P nf and P sf The relationship is P sf With P nf The ratio of represents the listening efficiency of the source semantic node for the stream f and the probability of data correctness. The basic framework of the semantic node intermediate state analysis model is characterized by the effectiveness of source semantic node listening.

[0084] Step S7202: Based on the particle swarm optimization algorithm, the maximum number of RTSs initiated by the network before the data packet is discarded is set to u r ,based on Figure 4 Construct the state transition process of the intermediate state analysis model of the semantic node.

[0085] Step S73: According to the obtained semantic node intermediate state analysis model, a collaborative group of a source semantic node set and a target semantic node set is established to perform data collaborative perception evolution modeling to obtain a system perception strategy based on precise semantic analysis;

[0086] Step S73 further comprises:

[0087] Step S7301: Set Y t Indicates the probability of a hidden node appearing each time a semantic node initiates RTS, 1-Y t It indicates the probability of the data packet achieving normal addressing each time RTS is initiated. Based on the semantic node intermediate state analysis model, the mapping relationship between the source semantic node set and the destination semantic node set is abstracted, and the nonlinear mapping between sets is used to construct the coordination group.

[0088] Step S7302: Utilize the vector characteristics and spatiotemporal correlation of the collaborative group to control the number of times each functional module of the semantic network initiates RTS and fails within a reasonable range, and obtain a system perception strategy based on precise semantic analysis based on the performance requirements of multi-information fusion perception.

[0089] In this embodiment, the performance evaluation is mainly based on the perception accuracy and the node connection capability, and the obtained perception strategy based on precise semantic analysis is the optimized algorithm architecture.

[0090] Step S74: According to the obtained precise semantic analysis system perception strategy, based on the language generation mechanism, dynamically control the channel state of the source semantic node before initiating the RTS, dynamically adjust the semantic relevance and spatiotemporal relevance of the destination semantic node, and ensure the fusion performance of multi-source language generation requirements;

[0091] This embodiment mainly implements adjustment and control based on the cohesiveness of natural language information, the relevance of words, and the dynamic mapping of grammar.

[0092] Step S75: Complete the fusion of language generation requirements with high correlation and strong spatiotemporal correlation from the perspective of fusion of multi-source similar information.

[0093] Step S8: Under static conditions, the data management convergence factor based on the language utility function is used; under dynamic conditions, the data management convergence factor based on the language utility function and the particle swarm optimization feedback parameter are combined to perform language processing and generate task processing and resource redistribution. The language processing of the intelligent manufacturing system based on collaborative perception is realized.

[0094] Specifically, this step includes:

[0095] Step S81: Based on the characteristic that the mathematical space R(N) covered by the present invention is a local compact space, the perception strategy model of the precise semantic analysis system based on static conditions and dynamic conditions is optimized respectively, and a solution of a multi-objective optimization genetic algorithm is designed according to the characteristic of the local compact space;

[0096] Step S82: Based on Figure 4 , let ∪A=N, for any n∈N, we can define Fn={1,2,…,n} and Un={{A}∪(A / Fn):A∈A}∪{{x}:x∈Fn}, then Un is an open cover of R(N). Based on this feature, the adjustability of semantic nodes is used to improve the coverage quality of language sensitive area monitoring, and the characteristics of language utility function are used to achieve collaboration and adjustment between semantic nodes under energy-constrained conditions;

[0097] Step S82 further comprises:

[0098] Step S8201: For any x∈R(N), when x∈Fn, st(x,Un)={x}, and x∈A, st(x,Un)={x}∪(x / Fn), so {Un} is the expansion of R(N). Let K be a compact subspace of R(N). Since A is a closed discrete subspace of R(N), K∩A is a finite set, so K is a countable set of R(N) and K is measurable, so K has a countable neighborhood basis in R(N). Based on this feature, the adjustability of semantic nodes is achieved and the coverage quality of language sensitive area monitoring is improved.

[0099] Step S8202: Assume that N is a countable dense subset of R(N), then R(N) is a separable space. If R(N) has a point countable basis, then R(N) must have a countable basis, so A, the subspace of R(N), will have a countable basis, which contradicts the assumption that A is an uncountable closed discrete subspace of R(N), so R(N) does not have a point countable basis. Based on this feature, collaboration and adjustment between semantic nodes are achieved under energy-constrained conditions.

[0100] Step S83: The network topology of the intelligent manufacturing system designed in this embodiment is set to belong to the Gillman-Jerison space, and the probability lower bound of the network connectivity can be set by the Poisson Boolean model. Based on this characteristic, the maximum value of the semantic elements and the language processing limit value of the system are designed to be transmitted simultaneously;

[0101] Step S83 further comprises:

[0102] Step S8301: Set the probability of connecting semantic nodes located in the interval 0≤x≤D to P oD (x) = P o (x)+P o (Dx)-P o (x)P o (Dx), where The average connection probability of a semantic node occupying the language utility function is P oD The lower bound of depends on the worst state of the node position, that is, P oD + =inf 0≤x≤D {P oD (x)}=2P o (D / 2)-P o 2 (D / 2). Because P o The lower bound of (x) is So when 2r≤D≤4r, P o (D / 2) = 1-e -λr -λDe -λr / 2 and P oD The lower bound of P oD + =1-(λD / 2+1) 2 e -2λr Based on this feature, the data management convergence factor of the language utility function is used to improve the language processing convergence stability.

[0103] Step S8302: Set the average value n of the number of semantic nodes sent in a stable state y =b l n t , where b l is the network congestion coefficient. To ensure the normal initiation of RTS, the minimum space requirement is Among them, M c is the carrier sense capacity, which indicates the maximum number of RTSs that can be initiated without packet collision. Combined with the particle swarm optimization feedback parameter, we have nodes can successfully transmit data. RTS initiated simultaneously, of which Mb is the data transmission capacity. Based on this characteristic, the maximum value of the semantic elements transmitted simultaneously by the system and the language processing limit of the system are obtained.

[0104] Step S84: Set the system comprehensive packet loss rate to According to the maximum value of semantic elements transmitted simultaneously by the system and the language processing limit value, the data management convergence factor based on the language utility function and the particle swarm optimization feedback parameter are combined to realize language processing and generation task processing and resource redistribution to meet the language collaborative perception needs of all nodes in the system.

[0105] Based on the initiation characteristics of multi-source languages ​​and the sensitivity of data collaborative perception mechanism, the accurate description of language collaborative perception decision is achieved;

[0106] In this embodiment, the description is achieved based on the cohesion of multi-source natural language information, the relevance of words, and the dynamic mapping of grammar.

[0107] Step S85: Based on the accuracy description of language collaborative perception decision, language processing, task generation processing and resource reallocation processing are implemented to complete the language processing of the intelligent manufacturing system based on collaborative perception.

[0108] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0109] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0110] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0111] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0112] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made by those skilled in the art within the spirit and principle of the present invention without creative labor shall be included in the protection scope of the present invention.

Claims

1. A language processing method for an intelligent manufacturing system based on collaborative perception, characterized in that: The following steps are involved: Initialize language task data, including total task completion time, energy consumption and network load status; According to the initialized data, a cost function is constructed and optimized using the particle swarm optimization algorithm. Based on the mapping relationship between particles and the nonlinear characteristics of the position and velocity of each particle, the data set of the system language model is obtained. The collaborative group is established based on the nonlinear characteristics of the language data set, and the numerical characteristics of the collaborative group are set by using the stickiness and semantic relevance of the language to ensure the realization of the node's ability to perceive the language processing requirements individually; Based on the principle that vectors are orthogonal to each other and there is no direct mapping, the idea of ​​joint optimization of language attributes and collaborative group feature vectors is used to make joint judgments and estimates of data of related functional nodes within the group. Establish a system node collaborative coverage model and set constraints for the node collaborative coverage model; Based on the collaborative coverage model, a dynamic balancing scheme is used to improve the connection strategy between system language generation tasks; Configure the characteristics of system routing self-adjustment and self-configuration in the case of multi-source language fusion, and perform intra-cluster coverage control based on the characteristics; From the perspective of multi-source similar information fusion, the language generation requirements with high correlation and strong spatiotemporal correlation are integrated; Under static conditions, a data management convergence factor based on language utility function is adopted; under dynamic conditions, a combination of a data management convergence factor based on language utility function and a particle swarm optimization feedback parameter is adopted to carry out language processing, generate task processing and resource redistribution, and realize language processing of intelligent manufacturing system based on collaborative perception.

2. The language processing method of an intelligent manufacturing system based on collaborative perception as claimed in claim 1, characterized in that: The process of constructing the cost function includes: The cost function is constructed with the minimum system energy consumption as the objective function and the maintenance of network semantic node connectivity as the constraint condition.

3. The language processing method of an intelligent manufacturing system based on collaborative perception as claimed in claim 1, characterized in that: The process of establishing a system node collaborative coverage model and setting constraints of the node collaborative coverage model includes: Based on the requirements of system network logic mapping coverage and energy consumption minimization, on plane A containing n network semantic nodes n The spatial constraint mechanism is used, and the logical mapping radius of each network semantic node is d; the network nodes are on plane A n The uniform distribution is followed, and the total semantic interference of the network is reduced by increasing the plane.

4. The language processing method of an intelligent manufacturing system based on collaborative perception as claimed in claim 1, characterized in that: Based on the collaborative coverage model, the process of improving the connection strategy between system language generation tasks by using a dynamic balancing solution includes: Let l i ,l j ∈R 2 Represents the network semantic node v i ,v j The position of v i ≠v j ; Define network semantic node v i and v j The direct connectivity is d; when there is a non-empty set in the network node group P, set the network semantic node v i and v j There are multi-hop connections between i and v j The information between can be effectively addressed and transmitted through the node group P; given the number of nodes is n, let l scr ,l1,l2,...,l n-1 ∈R 2 Represent the source node and nodes v1, v2, ..., v respectively. -1 , and let V be the set of all network semantic nodes; set N(l i ) is l i The maximum number of nodes in the logical mapping range with d as the center and d as the mapping radius, and N(l i )={v z :v z ∈Vand||l i -l z ||≤d}; all network semantic nodes have the same logical mapping coverage radius and mapping correlation, and the adjustability of system nodes is used to improve the coverage quality of sensitive area monitoring, and to achieve collaboration and adjustment among nodes under energy-constrained conditions.

5. The language processing method of an intelligent manufacturing system based on collaborative perception as claimed in claim 1, characterized in that: The process of configuring the self-adjustment and self-configuration characteristics of system routing in the case of multi-source language fusion and performing the logical mapping coverage control within the cluster includes: When merging multiple source languages, without considering the semantic boundary effect, the third-party network node v k Able to communicate with network nodes v j Achieve direct interconnection but cannot communicate with network nodes v i The upper bound for achieving direct interconnection is A(v j )-[A(v i )∩A(v j )], thereby realizing the characteristics of self-adjustment and self-configuration of system routing; Set the existence positions to be l i ,l j ∈R 2 Network nodes v i and v j , and ||l i -l j ||=d; ignoring the additional space of the last network node, the lower bound of the spatial position of the network node is Let T src is the total number of transmissions required when using the multi-hop method under error-free conditions, assuming that n network nodes are randomly located in area A according to uniform distribution n Within, the logical mapping coverage radius is d. When a network node transmits T = k times to realize a single bit broadcast to other (n-1) network nodes, its probability upper bound is G(k,n)(d 2 / A n ) n-1 ,in and This enables logical mapping coverage control within the cluster.

6. The language processing method of an intelligent manufacturing system based on collaborative perception as claimed in claim 1, characterized in that: From the perspective of multi-source similar information fusion, the process of integrating language generation requirements with high correlation and strong spatiotemporal correlation includes: Based on the relationship between semantic nodes and the semantic stability of the system, the probability of RTS initiated by the semantic node of the intelligent manufacturing system network for flow f is defined as P nf , define the probability of successful data transmission for flow f as P sf , initialize the energy consumption and load status of semantic nodes; According to the initialized parameters, a cost function is constructed in the system perception strategy process based on precise semantic analysis, and the particle swarm optimization algorithm is used for optimization to obtain the intermediate state analysis model of the semantic node; According to the obtained semantic node intermediate state analysis model, a collaborative group of source semantic node set and target semantic node set is established to model the data collaborative perception evolution and obtain a system perception strategy based on precise semantic analysis. According to the obtained precise semantic analysis system perception strategy, based on the language generation mechanism, the channel status of the source semantic node before initiating RTS is dynamically controlled, and the semantic relevance and spatiotemporal correlation of the destination semantic node are dynamically adjusted to ensure the fusion performance of multi-source language generation requirements; From the perspective of multi-source similar information fusion, the language generation requirements with correlation greater than the set value and spatiotemporal correlation intensity exceeding the preset value are integrated.

7. The language processing method of an intelligent manufacturing system based on collaborative perception as claimed in claim 6 is characterized in that: The process of optimizing using the particle swarm optimization algorithm to obtain the intermediate state analysis model of the semantic node includes: Define the probability that the destination semantic node becomes a hidden node due to exceeding the listening range of the source semantic node as Y t , in the steady state Y t With P nf and P sf The relationship is P sf With P nf The ratio of represents the source semantic node's listening efficiency and data correctness probability for stream f, and the basic framework of the semantic node intermediate state analysis model is characterized by the source semantic node listening effectiveness; Based on the particle swarm optimization algorithm, the maximum number of RTS initiated by the network before the data packet is discarded is set to u r , construct the state transition process of the intermediate state analysis model of the semantic node.

8. The language processing method of an intelligent manufacturing system based on collaborative perception as claimed in claim 6, characterized in that: According to the obtained semantic node intermediate state analysis model, a collaborative group of a source semantic node set and a target semantic node set is established, and the process of data collaborative perception evolution modeling includes: Set Y t Indicates the probability of a hidden node appearing each time a semantic node initiates RTS, 1-Y t It indicates the probability of a data packet achieving normal addressing each time an RTS is initiated; based on the semantic node intermediate state analysis model, the mapping relationship between the source semantic node set and the destination semantic node set is abstracted, and the coordination group is constructed using the nonlinear mapping between the sets; By utilizing the vector characteristics and spatiotemporal correlation of the collaborative group, the number of times the semantic network initiates RTS and fails is controlled within a reasonable range. Based on the performance requirements of multi-information fusion perception, a system perception strategy based on precise semantic analysis is obtained.

9. The language processing method of an intelligent manufacturing system based on collaborative perception as claimed in claim 1, characterized in that: Under static conditions, a data management convergence factor based on language utility function is adopted; under dynamic conditions, a combination of a data management convergence factor based on language utility function and a particle swarm optimization feedback parameter is adopted to perform language processing and generation task processing and resource redistribution. The process includes: Taking advantage of the fact that mathematical space is a locally compact space, the perception strategy model of the precise semantic analysis system based on static conditions and dynamic conditions is optimized respectively, and a solution of a multi-objective optimization genetic algorithm is designed according to the characteristics of the locally compact space. The adjustability of semantic nodes is used to improve the coverage quality of language-sensitive area monitoring, and the characteristics of language utility functions are used to achieve collaboration and adjustment between semantic nodes under energy-constrained conditions; The probability lower bound of the network connectivity is set through the Poisson Boolean model, and the maximum value of semantic elements transmitted simultaneously by the system and the language processing limit are set; Set the system comprehensive packet loss rate, and perform language processing and generation task processing and resource redistribution based on the maximum value of semantic elements transmitted simultaneously by the system and the language processing limit value, based on the data management convergence factor of the language utility function and the particle swarm optimization feedback parameter, to meet the language collaborative perception needs of all nodes in the system; Based on the initiation characteristics of multi-source languages ​​and the sensitivity of data collaborative perception mechanisms, the accuracy of language collaborative perception decisions is described; Based on the accurate description of language collaborative perception decision-making, language processing, task generation processing and resource reallocation processing are realized to complete the language processing of the intelligent manufacturing system based on collaborative perception.

10. The language processing method of an intelligent manufacturing system based on collaborative perception as claimed in claim 9, characterized in that: The process of setting the probability lower bound of the network connectivity through the Poisson Boolean model and setting the maximum value of the semantic elements transmitted simultaneously by the system and the language processing limit value includes: Set the probability of connecting semantic nodes in the interval 0≤x≤D to P oD (x) = P o (x)+P o (Dx)-P o (x)P o (Dx), where The average connection probability of a semantic node occupying the language utility function is P oD The lower bound of depends on the worst state of the node position, that is, P oD + =inf 0≤x≤D {P oD (x)}=2P o (D / 2)-P o 2 (D / 2); because P o The lower bound of (x) is So when 2r≤D≤4r, P o (D / 2) = 1-e -λr -λDe -λr / 2 and P oD The lower bound of P oD + =1-(λD / 2+1) 2 e -2λr ; Based on the above characteristics, the data management convergence factor of the language utility function is used to improve the convergence stability of language processing; Set the average number of semantic nodes sent in a stable state n y =b l n t , where b l is the network congestion coefficient. To ensure the normal initiation of RTS, the minimum space requirement is Among them, M c is the carrier sense capacity, which indicates the maximum number of RTSs that can be initiated without packet collision. Combined with the particle swarm optimization feedback parameter, we have nodes can successfully transmit data. RTS initiated simultaneously, of which M b is the data transmission capacity. Based on the above characteristics, the maximum value of the semantic elements that the system can transmit simultaneously and the language processing limit of the system are obtained.

Images