Device protocol and data dynamic matching method

Abstract

The invention discloses an equipment protocol and data dynamic matching method, which comprises the following steps of: firstly, modeling equipment information which comprises basic equipment information and protocol information; secondly, after the equipment is accessed to the edge computing node, the edge computing node obtains the basic information of the equipment and uploads the basic information to a protocol matcher of the Internet of Things platform; then the protocol matcher constructs a similar device group and a protocol group by using a similarity calculation technology, the edge calculation node receives the protocol group, issues the protocol group to the field device, calculates a protocol matching rate, and obtains an optimal matching protocol according to the matching rate; and finally, the edge computing node feeds back the matching rate of the optimal matching protocol to a protocol matcher of the Internet of Things platform, and dynamically adjusts parameters required for computing the similarity of the basic information of the equipment. According to the invention, the best matching protocol type and the supported data are matched for the equipment with unknown protocol information, and the matching efficiency of the protocol and the data in the equipment access process is improved.

Description

Technical Field

[0001] The present invention relates to the field of Internet of Things, and in particular to a method for dynamically matching device protocols and data. Background Art

[0002] Devices are a critical source of IoT data. Connecting diverse devices to the network through edge computing nodes and utilizing various protocols, enabling interoperability of device data, is driving the development of the IoT. However, with the deployment of massive heterogeneous devices, the number of connected devices is exploding. Faced with a vast array of diverse protocols, and in scenarios where device protocol information is unknown, efficiently generating optimally matched protocols for devices and rapidly connecting them to the network is crucial to the development of the IoT. Existing dynamic device onboarding methods rely on professionals to manually develop and select device protocols, which can be inefficient and costly when the number of protocols is large. Existing plug-and-play device onboarding methods, relying on standardized protocols, are unsuitable for scenarios with diverse protocols. While existing device protocol generation methods can predict protocol types, they cannot match device data or update parameters during the protocol matching process, resulting in low protocol matching rates. Furthermore, the inability to dynamically match access data necessitates manual configuration, resulting in low device onboarding efficiency. Summary of the Invention

[0003] To address the shortcomings of existing methods, the present invention proposes a method for dynamically matching device protocols and data. The system architecture of this method includes an IoT platform, edge computing nodes, and devices. The IoT platform includes a device information library, a protocol matcher, and a device protocol library. Edge computing nodes are deployed at the edge of the network, close to IoT devices, and connected to IoT devices via an on-site digital network (e.g., IP, RS485, etc.). The edge computing nodes and the IoT platform can communicate via wired or wireless networks.

[0004] The following details the device protocol and data dynamic matching:

[0005] Step 1: Model the device information. Device information includes basic device information and protocol information. Protocol information includes the protocol type and the data supported by the device. Device information is stored in the device information library of the IoT platform.

[0006] Step 2: After the device is connected to the edge computing node, the edge computing node obtains the basic information of the device and uploads the basic information of the device to the protocol matcher of the IoT platform.

[0007] Step 3: The protocol matcher uses similarity calculation techniques, such as edit distance, based on the device's basic information to obtain similarity calculation parameters. It then calculates the similarity between the access device's basic information and the basic information of all devices in the device information database. Based on this similarity, it constructs a group of similar devices. The matcher obtains the identifiers of the protocols supported by each device in the group and extracts protocols from the protocol database based on the protocol identifiers to construct a protocol group. This protocol group is then distributed to the edge computing node.

[0008] Step 4: The edge computing node receives the protocol group, obtains each protocol in the protocol group, and sends the identity verification message and data verification message in each protocol to the field device. The field device responds with an ACK and data message. The edge computing node calculates the protocol matching rate based on the response message and determines the best matching protocol based on the matching rate.

[0009] Step 5: The edge computing node feeds the matching rate of the best matching protocol back to the protocol matcher on the IoT platform. The optimizer in the protocol matcher dynamically adjusts the parameters required for calculating the similarity of basic device information based on the reported protocol matching rate. The protocol matcher consists of two parts: the inner layer is responsible for creating the protocol group and calculating the weighted penalty function, while the outer layer uses the optimizer to update the similarity parameters and construct similar device groups.

[0010] Beneficial effects of the present invention:

[0011] The present invention can match the best matching protocol type and supported data for devices with unknown protocol information based on the basic information of the device and using the existing device information database, thus eliminating the need for manual protocol selection and data configuration and improving the matching efficiency of protocols and data during device access. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. The drawings described below are some embodiments of the present invention. Those skilled in the art can also derive other drawings based on these drawings without inventive effort.

[0013] Figure 1 The device protocol dynamic matching framework provided by the present invention;

[0014] Figure 2 The device protocol matching network diagram provided by the present invention;

[0015] Figure 3 This is the expression form of the device information and protocol provided by the present invention. DETAILED DESCRIPTION

[0016] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0017] In this implementation, the edge computing node is an A100 with an ARM-based CPU, eight RS485 interfaces, and one CAN interface. It communicates with the IoT platform via 4G wireless communication. The device connects to the edge computing node via the RS485 interface. The IoT platform is deployed on Alibaba Cloud, where the server configuration includes a quad-core CPU, a 256GB hard drive, 64GB of RAM, 16M bandwidth, and a 2.3GHz clock speed.

[0018] To better describe the dynamic matching of device protocols and data, the following describes an implementation of the invention:

[0019] Step 1: Modeling equipment information and building equipment information database, such as Figure 1 shown.

[0020] The device information is defined as d = [i, P], where i represents the basic information of the device and P represents the protocol type and basic data set supported by the device. The basic information is defined as Where β is the size of the basic information vector. P is defined as P = [P id ,b], where P id represents the protocol identifier, and b represents the device data set obtained through P. Further, the device data set is defined as Where λ is the size of b.

[0021] Device information is expressed in the form of knowledge graph, such as Figure 3As shown in Figure 2. The core of the device ontology is the "Device" class. The "Device" class contains three object properties: hasBasicData, hasProtocol, and hasBasicInformation, indicating "basic data," "protocol," and "basic information," respectively. The Basic Information class contains two data properties: hasManu and hasModel, indicating "manufacturer" and "model," respectively. Because "basic data" is obtained through the protocol, its relationship with the "protocol" is obtains, meaning "obtain." The "Basic Data" class's data properties include hasName, hasUnit, hasType, and hasDesc, indicating that a data item has name, unit, type, and description information, respectively. Furthermore, "Basic Data" has an object property called hasParser, indicating "has a parser," responsible for obtaining the device's basic data according to the protocol. The "Protocol" class has both object and data properties. Object properties include hasVerification, required for protocol verification, and obtains, indicating "has verification" and "obtained data," respectively. Data properties include hasPID and hasDesc, indicating the presence of the protocol identifier and description, respectively. "Verification" has an object attribute hasProbe and a data attribute hasACK. hasProbe identifies the generation of a protocol probe packet, and hasACK represents the validity of the protocol.

[0022] To calculate the basic information similarity between two devices, the two devices d n With device d m The similarity between them is defined as:

[0023]

[0024] Where μ j Indicates device d n and equipment d m The word2vector similarity of the jth element in the basic information, w represents the weight vector, ε m The value of is proportional to the similarity between devices. A similarity of 0 indicates no similarity, and a similarity of 1 indicates complete similarity.

[0025] Step 2: Obtain basic device information.

[0026] Devices are connected to edge computing nodes via the RS485 bus. On-site staff enter basic device information or automatically retrieve it from the device. The edge computing node sends this basic device information to the IoT platform via the 4G network.

[0027] Step 3: Dynamically match device protocols.

[0028] After receiving the basic information of the device, the protocol matcher of the IoT platform converts the basic information of the device into the form of a knowledge graph. According to the basic information of the device, the required parameters are selected and the similar device group is calculated according to the similarity definition in step 1. The top 3 devices are sorted by the basic similarity with the access device to form a similar device group. Based on this similar device group, a protocol group P = [p1, p2, ..., p M ].

[0029] Step 4: Protocol verification.

[0030] The protocol group P is sent to the edge computing unit for verification. The edge computing unit parses P and obtains each protocol in P. In each protocol verification process, according to the definition of each protocol, according to P id The protocol matching rate is calculated by adding b. The protocol matching rate of the access device is defined as:

[0031]

[0032] Where λ * Represents the actual number of basic data of the access device, and λ represents the number of basic data matched by the access device. The device sends a protocol identification message to the device, and the device feedback ACK indicates that P id Correct, otherwise P id Incorrect. If P id Incorrect, the value of R is 0. If P id Correct, then the value of R is affected by two factors: P id The judgment value of and the judgment value of b. At this time, P id The judgment value of is fixed at 0.5, while the judgment value of b needs to be dynamically calculated. If the value of R is between 0.5 and 1, it means that P id Correct, but not all basic device data was found. According to the definition of protocol matching rate, the protocol matching rate in the edge computing node's protocol group is the highest, and the best matching protocol is the one with the highest matching rate. The edge computing node feeds the best matching protocol and matching rate back to the protocol matcher on the IoT platform.

[0033] Step 5: The protocol matcher updates the protocol matching parameters.

[0034] The network of the protocol matcher is as follows Figure 2 As shown in Figure 1, in order to maximize the matching rate of the best matching protocol, it is necessary to dynamically adjust the weight parameters required for the basic information similarity calculation. To determine the optimal weight w strategy, the present invention designs a penalty mechanism to update w according to the value function of w. The value function of w is defined as:

[0035] V(w)=μR,α∈[0,1]

[0036] Where μ is the weight of R. The penalty function is defined as:

[0037] P(w)=α(V MAX -V(w)),α∈[0,1]

[0038] Where V MAX Indicates the current maximum value of the w value function during the parameter adjustment process, and α is the penalty factor. In the process of finding the optimal w strategy, the update function of the element value in the weight vector is defined as:

[0039] w j* =w j +γ[-P(w)] + ,w j ∈w

[0040] Where w j* is the updated weight value, γ is the weight update discount factor, [M] + =max{0,M}.

[0041] During the update process of w, a single element w is selected and optimized in each iteration. j , with the remaining elements remaining unchanged. For each updated w, a similar device group and protocol group are dynamically constructed. The best matching protocol within this group is found, and its matching rate is calculated until a protocol with a matching rate that meets the protocol matching requirements is found. With iterative updates, w gradually approaches the global optimum, and V(w) approaches the global maximum, at which point the matching rate for the matching protocol is the highest.

[0042] Example:

[0043] In the present invention, two devices are taken as an example to illustrate the protocol matching process. The model of the edge computing node is A1000, and the models of the two connected devices are: ZXDU68B201 and ZXDU68S601, which represent the direct use of the protocol and dynamic matching respectively. The basic information (manufacturer, model and type) of the two devices are [ZTE, ZXDU68B201, UPS] and [ZTE, ZXDU68S601, UPS], respectively. Both devices are connected to the edge computing node via the RS485 bus. The device information in the device information library is shown in Table 1, where the manufacturer represents the manufacturer of the device, the model represents the model of the device, and the type represents the device type to which the device belongs.

[0044] Table 1 Device information in the device library

[0045]

[0046]

[0047] Scenario 1: Direct protocol usage. For the access device ZXDU68B201, its basic information similarity is compared with devices in the device information database. The highest similarity value is 1.0, indicating that the protocol for this device is already in the protocol database. Dynamic protocol construction is not required, and the protocol corresponding to the device with a similarity of 1 is directly used.

[0048] Scenario 2: Dynamic protocol matching. The access device ZXDU68S601 performs basic information similarity calculations with all devices in the device information database. The initial weights used in the similarity calculation are [1, 1, 1]. The top three similar device groups are shown in Table 2. The meanings of manufacturer, model, and type are the same as in Table 1.

[0049] Table 2 Similar device groups of access device ZXDU68S601

[0050] Manufacturer model type ZTE ZXDU68B201 ups ZTE ZXDU68W301 ups ZTE ZXDU68B301 ups

[0051] The protocols and basic data corresponding to the similar device group of the access device ZXDU68S601 are shown in Table 3, where "protocol" represents the protocol number, "device" represents the expression of the device, and "protocol identifier" is P id "Device data" is the basic data of the device.

[0052] Table 3 Protocols and basic data corresponding to the similar device group of device ZXDU68S601

[0053] protocol equipment Protocol Identifier Device data Protocol 1 ZXDU68B201 YDT / 1363 Ia,IF,LA,LB,TL,TC Protocol 2 ZXDU68W301 YDT / 1363 Ia,Ua,Ub,Uc,LA,LB,TD Protocol 3 ZXDU68B301 Modbus RTU Ib,Ic,Ua,IF,LB,TM,TX

[0054] This protocol group is distributed to the edge computing node. According to the matching rate calculation formula, the matching rates for Protocols 1, 2, and 3 in the protocol group are 0.83, 0.87, and 0.93, respectively. Therefore, Protocol 3 is the best match. Based on the value function of the weight vector, the matching rate weight is set to 0.5, and the calculated value of the weight is 0.465. Based on the definition of the penalty function, the penalty factor is set to 0.2, and the calculated value of the weight is -0.093. Based on the weight vector update function, the updated weight vector is [1.093, 1, 1], and the maximum value of the weight vector is updated to 0.465. This process continues iteratively until the highest matching rate is found.

Claims

1. A method for dynamically matching device protocols and data, characterized in that: The following steps are involved: Step 1: Model the device information, which includes basic device information and protocol information; Step 2: After the device is connected to the edge computing node, the edge computing node obtains the basic information of the device and uploads it to the protocol matcher of the IoT platform; Step 3: The protocol matcher uses similarity calculation technology to build similar device groups and protocol groups based on the basic information of the devices; Step 4: The edge computing node receives the protocol group, sends it to the on-site device, calculates the protocol matching rate, and obtains the best matching protocol based on the matching rate; Step 5: The edge computing node feeds back the matching rate of the best matching protocol to the protocol matcher. The optimizer in the protocol matcher dynamically adjusts the parameters required for calculating the similarity of the basic information of the device based on the protocol matching rate.

2. A device protocol and data dynamic matching method according to claim 1, characterized in that: The protocol information includes the protocol type and the data supported by the device. The device information is stored in the device information library of the Internet of Things platform.

3. A device protocol and data dynamic matching method according to claim 2, characterized in that: The step 3 is specifically implemented as follows: the protocol matcher uses similarity calculation technology to obtain similarity calculation parameters based on the basic information of the device, calculates the similarity between the basic information of the access device and the basic information of all devices in the device information database, and constructs a similar device group based on the basic information similarity; The matcher obtains the identifier of the protocol supported by each device in the similar device group, extracts the protocol from the protocol library according to the protocol identifier, builds a protocol group, and sends the protocol group to the edge computing node.

4. A device protocol and data dynamic matching method according to claim 3, characterized in that: The calculation of the similarity between the basic information of the access device and the basic information of all devices in the device information database is specifically as follows: n With device d m The similarity between them is defined as: Where μ j Indicates device d n and equipment d m The word2vector similarity of the jth element in the basic information, w represents the weight vector, ε m The value of is proportional to the similarity between devices. A similarity of 0 indicates no similarity, and a similarity of 1 indicates complete similarity.

5. A device protocol and data dynamic matching method according to claim 4, characterized in that: Step 4 is specifically as follows: the edge computing node receives the protocol group, obtains each protocol in the protocol group, sends the identification verification message and data verification message in each protocol to the field device, the field device replies with ACK and data messages, and the edge computing node calculates the protocol matching rate based on the reply message, and obtains the best matching protocol based on the matching rate.

6. A device protocol and data dynamic matching method according to claim 5, characterized in that: The specific implementation process of step 4 is as follows: The protocol group P is sent to the edge computing unit for verification. The edge computing unit parses P and obtains the protocol types supported by the device and each protocol in the basic data set P; In each protocol verification process, according to the definition of each protocol, based on the protocol identifier P id The protocol matching rate is calculated with the device data set b. The protocol matching rate of the access device is defined as: Where λ * represents the actual number of basic data of the access device, and λ represents the number of basic data matched by the access device; the device sends a protocol identification message to the device, and the device feedback ACK indicates that P id Correct, otherwise P id Incorrect; if P id Incorrect, the value of R is 0; if P id Correct, then the value of R is affected by two factors: P id The judgment value of and the judgment value of b, at this time, P id The judgment value of is fixed at 0.5, while the judgment value of b is dynamically calculated. If the value of R is between 0.5 and 1, it means that P id Correct, but not all basic data of the device is found; according to the definition of protocol matching rate, the matching rate of protocols in the edge computing node protocol group is the best matching protocol; the edge computing node feeds back the best matching protocol and matching rate to the protocol matcher on the IoT platform.

7. A device protocol and data dynamic matching method according to claim 6, characterized in that: The protocol matcher includes an inner layer and an outer layer. The inner layer is responsible for creating a protocol group and calculating a weight penalty function, while the outer layer is responsible for using an optimizer to update similarity parameters and build a similar device group.

8. A method for dynamically matching device protocols and data according to any one of claims 1 to 7, characterized in that: The parameters required for the dynamic adjustment of the device basic information similarity calculation are as follows: Design a penalty mechanism to update w according to the value function of the weight vector w. The value function of w is defined as: V(w)=μR,α∈[0,1] Where μ is the weight of R and the penalty function is defined as: P(w)=α(V MAX -V(w)),α∈[0,1] Where V MAX represents the current maximum value of the w value function during the parameter adjustment process, α is the penalty factor, and the update function of the element value in the weight vector during the optimal w strategy search process is defined as: w j* =w j +γ[-P(w)] + ,w j ∈w Where w j* is the updated weight value, γ is the weight update discount factor, [M] + =max{0,M}; During the update process of w, a single element w is selected and optimized in each iteration. j , and the other elements remain unchanged; for each updated w, dynamically construct similar device groups and protocol groups, find the best matching protocol in the protocol group, and calculate its matching rate until the protocol with the matching rate meets the protocol matching requirements. With iterative updates, w gradually approaches the global optimum, and V(w) approaches the global maximum value. At this time, the matching rate of the matched protocol is the highest.

Images