A robot rule real-time computing method and device based on cloud edge computing

Abstract

The application discloses a kind of robot rule real-time computing method and device based on cloud edge computing, comprising: visualization module is used to provide the editing function of rule and event, and edited rule data and event data are stored in storage module, also for showing calculation result and event sending result;Storage module is used to store rule data and event data;Calculation module is used to read rule data from storage module, and rule calculation is carried out by real-time acquisition feature data to obtain calculation result;Event module is used to read event data from storage module, and determine the sending channel of event, when obtaining rule calculation completion and reaching the condition of sending event, event is sent according to the sending channel of event by calculation module trigger, and the event carries rule calculation result.The method and device can realize the efficient transmission and sending of robot overall system, realize the monitoring of real-time robot running state and rule, realize the efficient information processing of robot.

Description

Technical Field

[0001] This invention relates to the field of robot automation control technology, specifically to a method and apparatus for real-time calculation of robot rules based on cloud-edge-device computing. Background Technology

[0002] A robot is an intelligent machine capable of semi-autonomous or fully autonomous operation. Through programming and automatic control, robots can perform tasks such as manual labor or movement, and are widely used in the medical, educational, subway inspection, and various aspects of production and daily life.

[0003] Robot behavior control refers to controlling a robot to make different reactions or actions through computer programs. Robot behavior control requires determining the robot's next action based on environmental information it perceives; these actions include movement, grasping objects, etc. The development of robot behavior control is moving towards greater flexibility and efficiency. Currently, some robots can make autonomous decisions and perform tasks in complex environments. For example, in the medical field, robots can perform surgical procedures or assist doctors in diagnosing diseases.

[0004] In the field of robot behavior control, a significant development in recent years has been deep reinforcement learning. Deep reinforcement learning refers to the use of deep neural networks combined with reinforcement learning algorithms to control intelligent robots. This method helps robots better adapt to complex environments and perform various tasks more efficiently.

[0005] Currently, robot behavior control largely relies on rule constraints. However, robot behavior is guided by various dynamically changing environmental, state, and logical control factors, leading to redundant and complex systems and repetitive robot design logic. Therefore, for a robot design system, it is necessary to design a robot functional subsystem that can operate independently of the robot decision-making system to address these issues. The advantage of this method lies in its ability to express complex rules through pseudocode, providing real-time, responsive rule computation, a long-term rule computation recording system, and an online tracking system.

[0006] A robot system is a complex engineering project involving multiple functional modules, each of which involves condition and threshold judgments and the selection of logical branches. Currently, rule calculation is handled independently by each module. However, based on the current development process of the robot cloud brain, this approach has the following problems: scattered and unconstrained calculations; permission and security risks; a large number of rules being calculated repeatedly; and rule calculations being less intuitive and making it difficult to track problems. Summary of the Invention

[0007] The purpose of this invention is to provide a method and apparatus for real-time calculation of robot rules based on cloud-edge computing, which can realize efficient transmission and sending of the entire robot system, real-time monitoring of robot operation status and rules, and efficient information processing of the robot.

[0008] To achieve the above-mentioned objectives, the embodiments also provide a real-time robot rule calculation device based on cloud-edge computing, including a visualization module, a calculation module, an event module, and a storage module;

[0009] The visualization module is used to provide editing functions for rules and events, and to store the edited rule data and event data in the storage module. It is also used to display the rule calculation results and event sending results.

[0010] The storage module is used to store rule data, event data, and rule calculation results;

[0011] The calculation module is used to read rule data from the storage module and perform rule calculations using real-time acquired feature data to obtain rule calculation results;

[0012] The event module is used to read event data from the storage module and determine the event sending channel. When the rule calculation is completed and the conditions for sending the event are met, the event is sent according to the event sending channel after being triggered by the calculation module. The event carries the rule calculation result.

[0013] Preferably, the event is a rule notification mechanism, where each event is bound to a rule, and a rule is bound to at least one event. Events are sent out after being triggered by a rule, according to a default binding order. Each event is defined as including an event name, a data table storing the robot's real-time data that triggered the rule, and a sending channel.

[0014] Preferably, the event is transmitted via a channel-based event transmission mechanism, wherein the event transmission mechanism includes:

[0015] Event queues triggered by the same rule can only be transmitted in one channel. External robots correspond to only one channel and listen for transmitted events on the corresponding channel.

[0016] During transmission, the event ID serves as a unique rule identifier, consisting of the external robot and a rule summary, and also includes event parameters that carry the rule calculation results.

[0017] Preferably, the rule data includes a rule tree list and a rule node list, wherein the rule tree is a nested structure, indicating that the current rule applies other rules, and the rule tree has logical symbols to represent the combinational logic of the rule calculation results;

[0018] Each rule tree also has rule nodes, each rule node corresponds to a rule and is used to store the rule's calculation expression. Each rule is given a unique name and ID, and the IDs of different rules are connected by logical symbols to form a rule tree.

[0019] Preferably, the robot rule real-time calculation device is implemented based on a rule engine, which includes a service interface layer, a rule and event map, a database, a management module, a scheduler, a thread pool, and a communication protocol module.

[0020] The service interface layer is used for front-end calls or proactive calls, including: querying, modifying, creating, and deleting events and rules, proactively triggering rule calculation, obtaining real-time rule calculation results, obtaining event sending records, and switching rule calculation status;

[0021] The rules and event maps contain rule and event instances in memory and reference each other.

[0022] The database is used for long-term storage of rules, events, logs, and computational backups;

[0023] The management module is used to provide access control constraints, control the execution and pause of rules in the engine, and add, delete, modify, and query rules and events;

[0024] The scheduler is used to determine when to calculate the scheduling rules;

[0025] The thread pool is used to abstract CPU computing resources and perform rule calculations;

[0026] The communication protocol module is used to send the bound events to the user or to synchronize the rule calculation results to the feature management platform.

[0027] Preferably, the process of rule calculation and event invocation notification implemented by the real-time computing device based on the rule engine includes:

[0028] Event entry: Register events according to the event definition. Registered events are saved to the database of the storage module and form an event queue.

[0029] Entry rules: Registration rules are defined according to the rules, and the registered rules are saved to the database of the storage module;

[0030] The rules engine monitors feature data updates from the feature platform;

[0031] Rules are retrieved sequentially from the rule queue according to the rule's polling cycle, or rules are retrieved from the rule queue when monitoring feature data is updated;

[0032] The obtained rules are judged, and when reading feature data from the feature platform, the rules are calculated to obtain the calculation result, and the calculation result is stored in the storage module or feature platform;

[0033] When a rule is bound to an event, the data table is packaged and the event is added to the event queue after the triggering conditions of the bound rule are met.

[0034] After obtaining the time from the event queue, the listener executes the response strategy corresponding to the event rule.

[0035] Preferably, during rule calculation, the rules are grouped according to different rule training times, and a timer is set for each group. When the timer expires, the real-time update calculation of a group of rules is triggered.

[0036] Preferably, for rule calculation, under a limited depth, a depth-first traversal method is used to traverse and transform the rule into a rule tree and record it.

[0037] Preferably, the rule tree is stored as a JSON data structure.

[0038] To achieve the above-mentioned objectives, the embodiments also provide a real-time robot rule calculation method based on cloud-edge-device computing. The method employs a real-time robot rule calculation device based on cloud-edge-device computing and includes the following steps:

[0039] The visualization module allows users to edit rules and events, store the edited rule and event data in the storage module, and also displays the calculation results and event sending results.

[0040] Use a storage module to store rule data and event data;

[0041] The calculation module reads rule data from the storage module and performs rule calculations using real-time acquired feature data to obtain the calculation results;

[0042] The event module reads event data from the storage module and determines the event sending channel. When the rule calculation is completed and the conditions for sending the event are met, the calculation module triggers the event and sends the event according to the event sending channel. The event carries the rule calculation result.

[0043] Compared with the prior art, the beneficial effects of the present invention include at least the following:

[0044] Based on defined rules and events, a rule calculation system that is real-time adjustable, visible, and fast-responding, as well as an autonomous and usable event broadcasting channel, separates the robot's complex perception, judgment, and behavior tree branching requirements from the existing robot control and decision-making theory system, thereby achieving efficient and intelligent system decision-making. Attached Figure Description

[0045] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0046] Figure 1 This is a schematic diagram of the structure of the real-time robot rule computing device based on cloud-edge computing provided in the embodiment;

[0047] Figure 2 This is a functional diagram of the rule engine provided in the embodiment;

[0048] Figure 3 This is a schematic diagram of the event definition provided in the embodiment;

[0049] Figure 4 This is a schematic diagram of the rule definition provided in the embodiment;

[0050] Figure 5 This is a flowchart of the rule engine execution process provided in the embodiment;

[0051] Figure 6 This is the rule calculation process provided in the implementation example;

[0052] Figure 7 This is the calculation process of the composite rules provided in the embodiment;

[0053] Figure 8 This is a flowchart of a real-time robot rule calculation method based on cloud-edge computing provided in the embodiment. Detailed Implementation

[0054] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and do not limit the scope of protection of this invention.

[0055] The inventive concept of this invention is as follows: Addressing the technical problems in existing robot systems where rule calculations for each module are processed separately, resulting in scattered and unconstrained calculations, insufficient intuitiveness, and difficulty in problem tracking, this invention provides a real-time robot rule calculation method and device based on cloud-edge-device computing. This method separates the complex perception, judgment, and behavior tree branching requirements of robots from existing robot control and decision-making theories, enabling a rule calculation system that is adjustable in real time, visible, and responsive, and providing an autonomous and usable event broadcast channel. It also enables rapid response to behavior tree decision-making branches, serving as a replacement for behavior tree data exchange backplanes.

[0056] Based on the above inventive concept, the embodiment provides a real-time robot rule calculation device based on cloud-edge-device computing, such as... Figure 1 As shown, it includes a visualization module, a calculation module, an event module, and a storage module. The visualization module provides online observation, configuration, and management functions for rules. Specifically, it provides editing functions for rules and events. Users can write, view, and modify rule data and event data through the visualization module, and store the edited rule data and event data in the storage module and / or memory cache. It is also used to display the rule calculation results and event sending results.

[0057] The storage module is used to store rule data and event data, as well as to store the real-time calculation results of rules for a certain period of time. It contains a database to store this data, and also stores the log data of the entire system.

[0058] The calculation module reads rule data from the storage module and performs rule calculations using real-time acquired feature data to obtain the result. Specifically, rule calculations can be performed periodically or at set intervals, or triggered by feature data updates.

[0059] The event module is used to read event data from the storage module and determine the event sending channel. When the rule calculation is completed and the conditions for sending the event are met, the calculation module triggers the event and sends the event to the rule listener according to the event sending channel. The event carries the rule calculation result.

[0060] The real-time robot rule calculation device provided in the above embodiments is implemented based on a rule engine. This rule engine can be developed using the Java language and uses the Spring framework as the network communication engine, such as... Figure 2 As shown, it includes a service interface layer, rules and event map, database, management module, scheduler, thread pool and communication protocol module.

[0061] The service interface layer provides a RESTful engine access mechanism for frontend or proactive invocation, including: querying, modifying, creating, and deleting events and rules; proactively triggering rule calculation; obtaining real-time rule calculation results; obtaining event sending records; and switching the rule calculation status, which includes running, paused, and invoked.

[0062] The rules and events map contains rule and event instances that reference each other. For example, the events triggered by rule A consist of e1, e2, and e3. The rules bound to event E consist of A, E, and F. The events that rule A is about to trigger are e1 and e3, and the packaging parameter is params1.

[0063] The database is used for long-term storage of rules, events, logs, and computation backups. Within the storage module's database, rules and events are stored using a defined data structure, which is directly saved in character mode. Each rule computation independently retains its computation parameters and status logs, saving them as local files. Computation backups are stored in memory, retaining the results of the last n rule computations for each rule, where n can be 10.

[0064] The management module provides access constraints, controls the execution and pausing of rules in the engine, and manages the creation, deletion, modification, and querying of rules and events. Specifically, it includes user permission constraints, ensuring that users can only modify rules and events they themselves create, and can only actively trigger rules they have created.

[0065] The scheduler is used to schedule the timing of rule calculations. Specifically, it is responsible for the timing of each rule calculation, which can be divided into trigger-based, polling-based, and dormant-based. During dormant-based calculations, the rule is not calculated under any circumstances. Trigger-based calculations calculate the rule once when the parameters related to the rule calculation change. Polling-based calculations trigger rule calculations periodically at regular intervals, with different polling times for each rule (which can be in the range of 5 milliseconds, 10 milliseconds, 20 milliseconds, 50 milliseconds, 100 milliseconds, or 1 second).

[0066] Thread pools are used to abstract CPU computing resources and perform rule-based calculations. In this example, by default, a computer with more than 20 core threads is used to support the calculations, and the thread pool occupies 50-80% of the computer's computing resources.

[0067] The communication protocol module is used to send bound events to the user or synchronize rule calculation results to the feature management platform. The communication protocol is a signaling message protocol, with a single data transmission size of less than 10KB. The communication protocol includes, but is not limited to, any IoT communication protocol and internet protocol, such as MQTT, RabbitMQ, and WebSocket. Parameter transmission in the communication protocol consists of robot ID, parameter ID, and parameter entity; event parameter transmission in the communication protocol consists of robot ID, event ID, event channel, and event entity.

[0068] In this embodiment, an event is defined for each robot. The format of this event includes an event name, a data table, and a transmission channel. Each event is a rule notification mechanism, carrying key information to the listener to inform them of the rule's execution status immediately. To facilitate quick use of events by rule engine users and avoid duplicate event creation, the rule engine provides several basic events (such as battery notifications and robot state transitions) that can be directly referenced. Events must be bound to rules to be triggered; that is, each event is bound to a rule, and each rule is bound to at least one event. Events are sent according to the default binding order after being triggered by a rule. The receiving end is provided with both a communication service that guarantees the event sending order and a service that does not guarantee the event sending order.

[0069] In this embodiment, events are sent via a channel-based event transmission mechanism. This mechanism includes: event queues triggered by the same rule can only be transmitted in one channel to avoid duplicate event reception caused by multiple channels; external robots correspond to only one channel and listen for transmitted events on that channel, ensuring efficient transmission under an isolated event transmission mechanism; during transmission, the event ID serves as a unique rule identifier, composed of the external robot and a rule summary; and it also includes event parameters (Event Content), representing the data carried when an event is generated and sent. Specifically, these event parameters include robot ID, event ID, event channel, and event entity, with the event entity including the rule calculation result.

[0070] Figure 3 An event definition diagram is provided. For example, an event ID that triggers an active welcome is defined as event_activive_welcome, the channel is / welcome, the feature list is humans, robotPos, and gazeNum, and the event entity is {type:'fast'}, indicating a fast welcome.

[0071] In this embodiment, the defined rules are stored using a data structure, which can be a JSON data structure, but is not limited to any data type capable of expressing a tree structure. In each rule data structure, the JSON parameters represent the rule ID, running status, and bound robot ID, used to distinguish different robots.

[0072] The rule data structure includes a list of rule trees (ruleTrees) and a list of rule nodes (condList), such as... Figure 4As shown, the rule tree is a nested structure, indicating that the current rule applies other rules. The rule tree has combinational logic representing the rule calculation results using logical symbols. Each rule tree also has rule nodes, each corresponding to a rule and used to store the rule's calculation expression. Each rule is assigned a unique name and ID, and the IDs of different rules are connected by logical symbols to form the rule tree.

[0073] The rule tree is stored in the database in JSON format for backup. When the rule engine starts, it reads the rule tree into memory and stores it as key-value pairs, forming a rule queue. Each rule is read according to the refresh frequency defined in the rule definition. The computation service traverses the rule tree to access each computation expression, calculates the result based on the expression, and then merges the results using logical symbols, passing the results up the tree level by level for rapid judgment and calculation. The computation expression includes conditional expressions used for conditional judgment during a specific traversal. Each rule has a robot list (robotIds) representing all robot IDs supported by the rule. During rule judgment, only parameters issued or rules triggered by robots in the list take effect.

[0074] To clearly illustrate how the rule engine operates, the embodiment also provides the process of rule calculation and event invocation notification, such as... Figure 5 As shown, it specifically includes:

[0075] (1) Entering events: Register events according to the event definition. For example, register a robot behavior tree event using the event ID event_btree. The registered events are saved to the database of the storage module and form an event queue.

[0076] (2) Entry rules: Register rules according to the rule definition. For example, the registration rule is rule_hello (say hello). The rule expression is designed to trigger the rule when the feature human.size (the number of people the robot sees) is greater than 1. The registered rules are saved to the database of the storage module.

[0077] (3) The rule engine listens for feature data updates from the feature platform. For example, if the rule engine receives a change in the feature human.size, it determines whether it is greater than 1. If the condition is not met, it skips the event; if the condition is met, it triggers the event_btree.

[0078] (4) Rules are extracted from the rule queue in sequence according to the rule polling cycle, or when the monitoring feature data is updated; this process means that the rule engine obtains each rule instance according to the rule calculation interval.

[0079] (5) Judge the obtained rules. If the rule is a leaf node, perform recursive rule calculation and merging. When reading feature data from the feature platform, perform rule calculation to obtain the calculation result and store the calculation result in the storage module or feature platform. For example, if the rule rule_hello is true, store it in the feature platform.

[0080] (6) When a rule is bound to an event, the data table is packaged and the event is added to the event queue after the triggering conditions of the bound rule are met;

[0081] The event is then added to the event queue, ready to be sent in real time on the channel / welcome;

[0082] (7) After obtaining the time from the event queue, the listening robot executes the response strategy of the corresponding rule for the event. For the rule rule_hello, if the condition is met, the event event_btree is triggered and received by the event processing module, and the robot says "hello".

[0083] It should be noted that the feature platform is an online data recording and caching service that provides real-time dynamic data access services.

[0084] In this embodiment, a method for calculating flow rules is also defined. Currently, the service uses single-machine computing, with a CPU-intensive server hosting the rule engine calculation service. This service creates a thread pool to calculate each rule independently, making full use of computing resources. Figure 6 As shown, rules are grouped according to different training times, and a timer is set for each group. When the timer expires, the real-time update calculation of a group of rules is triggered. For rule calculation, under a limited depth, a depth-first traversal is used to traverse the rule tree converted into JSON, and the results are recorded, as shown below. Figure 7 As shown.

[0085] Based on the same inventive concept, the embodiment also provides a real-time robot rule calculation method based on cloud-edge-device computing. The method employs the aforementioned real-time robot rule calculation device based on cloud-edge-device computing, such as... Figure 8 As shown, it includes the following steps:

[0086] Step 1: Use the visualization module to edit rules and events, and store the edited rule data and event data in the storage module. It also displays the calculation results and event sending results.

[0087] Step 2: Use the storage module to store rule data and event data;

[0088] Step 3: Use the calculation module to read the rule data from the storage module, and use the real-time acquired feature data to perform rule calculations to obtain the calculation results;

[0089] Step 4: Use the event module to read event data from the storage module and determine the event sending channel. When the rule calculation is completed and the conditions for sending the event are met, the event is sent according to the event sending channel after being triggered by the calculation module. The event carries the rule calculation result.

[0090] The above embodiments provide a real-time robot rule calculation method and device based on cloud-edge-device computing, which has autonomous decision-making and behavior tree, task management and scheduling, provides an event triggering mechanism, unified permission management, unified computing platform, efficient real-time computing processing capabilities, visualized result presentation and log tracking, and can realize efficient transmission and sending of the entire robot system, realize real-time monitoring of robot operation status and rules, and realize efficient information processing of the robot.

[0091] The specific embodiments described above illustrate the technical solution and beneficial effects of the present invention in detail. It should be understood that the above description is only the most preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, additions, and equivalent substitutions made within the scope of the principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A real-time robot rule calculation device based on cloud-edge-device computing, characterized in that, It includes a visualization module, a calculation module, an event module, and a storage module; The visualization module is used to provide editing functions for rules and events, and to store the edited rule data and event data in the storage module. It is also used to display the rule calculation results and event sending results. The storage module is used to store rule data, event data, and rule calculation results; wherein, the event is a rule notification mechanism, each event is bound to a rule, and a rule is bound to at least one event. Events are sent after being triggered by a rule according to the default binding order. Each event is defined as including an event name, a data table storing the real-time data of the robot that triggered the rule, and a sending channel. The calculation module is used to read rule data from the storage module and perform rule calculations using real-time acquired feature data to obtain rule calculation results; The event module is used to read event data from the storage module and determine the event sending channel. When the rule calculation is completed and the conditions for sending the event are met, it is triggered by the calculation module and sends the event according to the event sending channel. The event carries the rule calculation result. The robot rule real-time calculation device is implemented based on a rule engine, which includes a service interface layer, a rule and event map, a database, a management module, a scheduler, a thread pool, and a communication protocol module. The service interface layer is used for front-end calls or proactive calls, including: querying, modifying, creating, and deleting events and rules, proactively triggering rule calculation, obtaining real-time rule calculation results, obtaining event sending records, and switching rule calculation status; The rules and event maps contain rule and event instances in memory and reference each other. The database is used for long-term storage of rules, events, logs, and computational backups; The management module is used to provide access control constraints, control the execution and pause of rules in the engine, and add, delete, modify, and query rules and events; The scheduler is used to determine when to calculate the scheduling rules; The thread pool is used to abstract CPU computing resources and perform rule calculations; The communication protocol module is used to send the bound events to the user or to synchronize the rule calculation results to the feature management platform.

2. The real-time robot rule calculation device based on cloud-edge computing according to claim 1, characterized in that, The events are transmitted via a channel-based event transmission mechanism, wherein the event transmission mechanism includes: Event queues triggered by the same rule can only be transmitted in one channel. External robots correspond to only one channel and listen for transmitted events on the corresponding channel. During transmission, the event ID serves as a unique rule identifier, consisting of the external robot and a rule summary, and also includes event parameters that carry the rule calculation results.

3. The real-time robot rule calculation device based on cloud-edge computing according to claim 1, characterized in that, The rule data includes a list of rule trees and a list of rule nodes. The rule trees are nested, indicating that other rules are applied to the current rule. The rule trees have logical symbols to represent the combinational logic of the rule calculation results. Each rule tree also has rule nodes, each rule node corresponds to a rule and is used to store the rule's calculation expression. Each rule is given a unique name and ID, and the IDs of different rules are connected by logical symbols to form a rule tree.

4. The real-time robot rule calculation device based on cloud-edge computing according to claim 1, characterized in that, The process of rule calculation and event invocation notification implemented by the real-time computing device based on the rule engine includes: Event entry: Register events according to the event definition. Registered events are saved to the database of the storage module and form an event queue. Entry rules: Registration rules are defined according to the rules, and the registered rules are saved to the database of the storage module; The rules engine monitors feature data updates from the feature platform; Rules are retrieved sequentially from the rule queue according to the rule's polling cycle, or rules are retrieved from the rule queue when monitoring feature data is updated; The obtained rules are judged, and when reading feature data from the feature platform, the rules are calculated to obtain the calculation result, and the calculation result is stored in the storage module or feature platform; When a rule is bound to an event, the data table is packaged and the event is added to the event queue after the triggering conditions of the bound rule are met. After obtaining the time from the event queue, the listener executes the response strategy corresponding to the event rule.

5. The real-time robot rule calculation device based on cloud-edge computing according to claim 1, characterized in that, During rule calculation, rules are grouped according to different rule training times, and a timer is set for each group. When the timer expires, the real-time update calculation of a group of rules is triggered.

6. The real-time robot rule calculation device based on cloud-edge computing according to claim 1, characterized in that, For rule calculation, under a limited depth, a depth-first traversal is used to transform the rule tree and record the results.

7. The real-time robot rule calculation device based on cloud-edge computing according to claim 1, characterized in that, The rule tree is stored as a JSON data structure.

8. A real-time robot rule calculation method based on cloud-edge-device computing, characterized in that, The method employs the real-time robot rule calculation device based on cloud-edge computing as described in any one of claims 1-7, and includes the following steps: The visualization module allows users to edit rules and events, store the edited rule and event data in the storage module, and also displays the calculation results and event sending results. The storage module stores rule data and event data. The event is a rule notification mechanism. Each event is bound to a rule, and a rule is bound to at least one event. Events are sent after being triggered by a rule, according to the default binding order. Each event is defined as including an event name, a data table storing the real-time data of the robot that triggered the rule, and a sending channel. The calculation module reads rule data from the storage module and performs rule calculations using real-time acquired feature data to obtain the calculation results; The event module reads event data from the storage module and determines the event sending channel. When the rule calculation is completed and the conditions for sending the event are met, the calculation module triggers the event and sends the event according to the event sending channel. The event carries the rule calculation result.

Images