CN110866392A - System for strategy control and strategy control method thereof - Google Patents

Abstract

The invention provides a system for strategy control and a strategy control method using the system, wherein the system comprises a central control system, at least two robots and a scheduling strategy database, the central control system is provided with an accepting module, a calling module, a decision-making module and a feedback module, the accepting module receives inquiry contents, the calling module calls a preset scheduling strategy in the scheduling strategy database, the decision-making module makes a decision, the decision-making comprises the steps of selecting a conversation robot meeting the current conversation requirement from the at least two robots and selecting the optimal conversation result from the conversation results of the conversation robots; the communication module of the conversation robot is connected with the central control system and used for receiving inquiry content and giving a callback result; the feedback module feeds the optimal answer result back to the user; the invention realizes the flexible scheduling and the selection of the optimal answer for at least two robots by configuring the preset scheduling strategy, thereby achieving the technical effects of seamless switching user experience and convenient expansion.

Description

System for strategy control and strategy control method thereof

Technical Field

The invention mainly relates to the field of artificial intelligence natural language processing, in particular to a strategy control method for conversation sessions of at least two robots, which realizes flexible scheduling and optimal answer selection of a robot cluster.

Background

The robot technology is a strategic technology integrating multidisciplinary crossing of machinery, information, materials, biomedicine and the like, is one of the most active fields of advanced high-tech research at present, belongs to the strategic emerging industry of China, can not only deal with the problems of labor cost rise, population aging and the like, but also support and guide the development of related technologies and industries, and has very important significance for promoting the development of intelligent manufacturing equipment, enhancing the military national defense strength, improving the emergency treatment emergency capacity and developing medical rehabilitation equipment.

In the current stage of man-machine natural language interaction, an independent dialogue robot is generally adopted to realize single-scene interaction, the service fields and the service interaction flows of all robots are mutually independent, a user may need to jump among a plurality of robots back and forth in the process of receiving the robot service, and the user experience is hard and unfriendly.

The central control scheduling mode is that the robot selection strategy parameters are configured, and interface standard access is carried out on the robot group, so that scheduling selection and optimal answer giving of at least two robots become possible, meanwhile, the user experience is smooth and fluent, and the human-computer interaction experience is smooth.

At present, enterprises commonly use independent robots to realize a single scene service function, so that more and more robots are simultaneously operated in a system, a central control system capable of enabling multiple robot groups to cooperatively work is urgently needed, and an effective flow and a selection scheme can be provided to solve the problem that different robots dispute the same problem.

In the prior art, the answer robot is mostly directly allocated through semantic understanding of question sentences, the effect is good or bad, only one factor of a semantic understanding module is relied on, so that the answer accuracy is low, and the optimal selection is difficult to perform if the semantics of two robots are in conflict or difficult to distinguish.

The prior art chinese patent CN106782547A discloses a robot semantic recognition system based on speech recognition, which includes: the voice recognition unit is used for recognizing voice information of a user and converting the voice information into natural language to be recognized; the language receiving unit is used for receiving the natural language to be recognized; a semantic recognition unit for recognizing the natural language to be recognized received by the language receiving unit and feeding back a plurality of semantic recognition results associated with the natural language to be recognized; and the semantic recognition unit is used for confirming a final semantic recognition result from the plurality of semantic recognition results fed back by the semantic recognition unit according to the language habit of the user and the similarity of the natural language to be recognized. The language finally fed back by the technical scheme needs the user to confirm the final recognition result, and if the user does not confirm or feed back, the system cannot automatically perform subsequent processing.

The technology of the invention does not need the user to confirm again, the central control system can select the optimal answer call result to be displayed to the user according to the preset scheduling strategy, and in addition, the prior art is only a semantic recognition system, can not configure the scheduling strategy and can not reply to the inquiry action sent by the client.

Disclosure of Invention

The invention introduces a system for strategy control and a strategy control method using the system, and the specific technical scheme is as follows:

a system for policy control, comprising: the system comprises a central control system, at least two robots and a scheduling strategy database; the central control system is provided with a receiving module for receiving input inquiry content; the robot is provided with a communication module for receiving the inquiry content and giving a callback result corresponding to the inquiry content; the central control system is provided with a calling module and a decision module, the calling module is used for calling a preset scheduling strategy in a scheduling strategy database, the decision module is used for making a decision, and the decision comprises selecting a conversation robot meeting the current conversation requirement from the at least two robots and selecting an optimal conversation result from the conversation results of the conversation robots meeting the current conversation requirement; the preset scheduling strategy is configurable; the central control system is provided with a feedback module for optimizing the answer result. The user can input the inquiry content, and the central control system feeds back the optimal answer result to the user.

Further, the scheduling policy database includes a plurality of preset scheduling policies, and the preset scheduling policies for invoking the conversation robot include personalized information, conversation robot priority, intention, and context of the current conversation.

Further, the scheduling policy database comprises a plurality of preset scheduling policies, and the preset scheduling policies for retrieving the optimal answer result comprise the current state of the conversation robot, the confidence score of the answer and the intention.

Further, the conversation robot gives the callback result, including the confidence score, the current state and the intention of the callback result.

Furthermore, the at least two robots are in communication connection with the central control system and used for recording the real-time states of the at least two robots.

Further, the at least two robots are accessed through a standardized interface.

The invention also provides a strategy control method using the strategy control system, which comprises the following steps:

the receiving module receives inquiry content input by a user;

the central control system calls a preset scheduling strategy, wherein the preset scheduling strategy is a scheduling strategy for selecting the conversation robot;

aiming at the inquiry content input by a user, a central control system selects a conversation robot meeting the current conversation requirement from at least two robots according to a scheduling strategy of selecting the conversation robot;

sending inquiry content input by a user to the conversation robot;

the conversation robot receives the inquiry content and gives a callback result;

the central control system calls a preset scheduling strategy, wherein the preset scheduling strategy is a scheduling strategy for selecting an optimal answer result;

aiming at the answer results of the conversation robot, the central control system selects the optimal answer result from the answer results of the conversation robot according to the scheduling strategy for selecting the optimal answer result;

and the central control system feeds back the optimal answer result to the user.

Further, after the receiving module receives the inquiry content input by the user, the central control system calls a preset scheduling strategy, wherein the preset scheduling strategy comprises personalized information, the priority, the intention and the context of the current conversation of the conversation robot, and the central control system makes a decision for selecting the conversation robot from at least two robots according to the preset scheduling strategy.

Further, after receiving the inquiry content, the conversation robot selects whether to make a call back according to the current state, and if the call back is made, the communication module of the conversation robot feeds back the call back result to the central control system.

Further, after receiving the answer result of the conversation robot, the central control system calls a preset scheduling strategy, wherein the preset scheduling strategy comprises the current state of the conversation robot, the confidence score and the intention of the answer, and the central control system makes a decision for selecting the optimal answer result from the answer result of the conversation robot according to the preset scheduling strategy.

The invention has the beneficial effects that: the invention sets a configurable preset scheduling strategy function, the central control system selects at least two conversation robots to answer the call according to the preconfigured scheduling strategy, the central control system evaluates the answer results of the conversation robots again according to the preconfigured scheduling strategy, and selects the optimal answer result to display to the user; the invention realizes the flexible scheduling and the optimal selection of the callback result of the robot cluster by the configuration of the selection strategy of the robot and the configuration of the callback result selection strategy, thereby achieving the technical realization of seamless switching user experience and convenient expansion.

Drawings

FIG. 1 is a schematic block diagram of a system for policy control;

fig. 2 is an operational flow diagram of a policy control method.

Detailed Description

The present invention is described in detail below with reference to the attached drawings.

A first embodiment of the invention is shown in fig. 1, a system for policy control comprising: the system comprises a central control system 1, at least tworobots 6 and ascheduling strategy database 4; the central control system is provided with areceiving module 2 for receiving the inquiry content input by the user; therobot 6 has acommunication module 602 for receiving the inquiry content input by the user and giving a callback result corresponding to the inquiry content; the central control system is provided with acalling module 3 and adecision module 5, wherein thecalling module 3 is used for calling a preset scheduling strategy in ascheduling strategy database 4, and thedecision module 5 is used for making a decision, wherein the decision comprises selecting aconversation robot 601 meeting the current conversation requirement from at least tworobots 6 and selecting an optimal conversation result from the conversation results of the conversation robots meeting the current conversation requirement; the preset scheduling strategy is configurable; the central control system is provided with afeedback module 7 for feeding back the optimal answer result to the user.

Further, the scheduling policy database comprises a plurality of preset scheduling policies; the preset scheduling policy for invoking the conversation robot includes personalized information, conversation robot priority, intention, context of the current conversation.

Further, the preset scheduling policy for retrieving the optimal answer result includes the current state of theconversation robot 601, the confidence score of the answer, and the intention.

Further, theconversation robot 601 gives the answer result, including the confidence score, the current status, and the intention of the answer result.

The at least tworobots 6 are in communication connection with the central control system 1 and are used for recording the real-time states of the at least tworobots 6

Further, the at least tworobots 601 are accessed through a standardized interface.

A second embodiment of the invention is shown in fig. 1-2: a policy control method using the system of embodiment 1, comprising the steps of:

step S1: a receivingmodule 2 of the central control system receives inquiry content input by a user;

step S2: the central control system 1 calls a preset scheduling strategy in ascheduling strategy database 4 through acalling module 3, wherein the preset scheduling strategy is a scheduling strategy for selecting theconversation robot 601;

step S3: aiming at the inquiry content input by a user, a central control system selects one ormore conversation robots 601 meeting the current conversation requirements from at least tworobots 6 according to a scheduling strategy of selecting the conversation robots;

step S4: sending query content input by a user to aconversation robot 601;

further, the step of sending the inquiry content input by the user to the conversation robot may be set such that the central control system 1 sends the inquiry content of the user to theconversation robot 601, or may be set such that the user directly sends the inquiry content to theconversation robot 601;

step S5: theconversation robot 601 receives the inquiry content and gives a callback result;

further, thecommunication module 602 of the conversation robot receives the inquiry content and feeds back the answer result to the central control system 1;

further, thecommunication module 602 of the conversation robot may be configured to directly receive the query content input by the user;

further, after receiving the inquiry content, theconversation robot 601 selects whether to make a call back according to the current state, and if the call back is made, thecommunication module 602 of the conversation robot feeds back the call back result to the central control system 1;

step S6: after the central control system 1 receives the answer results of one or more conversation robots, the central control system 1 calls a preset scheduling strategy in ascheduling strategy database 4 through acalling module 3, wherein the preset scheduling strategy is a scheduling strategy for selecting the optimal answer result;

step S7: aiming at the answer results of the conversation robot, the central control system 1 selects the optimal answer result from the answer results of the conversation robot according to the scheduling strategy for selecting the optimal answer result;

step S8: the central control system 1 feeds back the optimal answer result to the user through afeedback module 7;

preferably, in step S3, after the receivingmodule 2 receives the query content input by the user, the central control system 1 invokes a preset scheduling policy, where the preset scheduling policy includes personalized information, a priority of the conversation robot, an intention, and a context of the current conversation, and the central control system 1 makes a decision to select a conversation robot satisfying a current conversation requirement from at least two robots according to the preset scheduling policy.

The preset scheduling policy for selecting the dialogue robot can be set according to the following rules:

1. and scheduling according to the personalized information of the user, such as gender, position information, access control mode and the like.

2. And scheduling the priority level of the currently executed task.

3. Scheduling according to intent.

Preferably, in step S6, after the central control system 1 receives the answer results of the conversation robot, the central control system 1 invokes a preset scheduling policy including the current state of the conversation robot, the confidence score of the answer, and the intention, and the central control system 1 makes a decision to select the optimal answer result from among the answer results of the conversation robot according to the preset scheduling policy.

The preset scheduling policy for selecting the optimal answer selection policy result may be set according to the following rules:

1. and determining according to the state of the conversation robot of the answer, selecting a normal state and not selecting an abnormal state.

2. And taking the highest score or limiting the range according to the score of the returned call.

3. And (5) directly answering according to intention matching.

The preset scheduling strategy in the invention can be defined or developed by a user in a certain range, thereby ensuring the accuracy to the maximum extent, and simultaneously providing convenience and expandability for the standardized interface access to the access of other subsequent robots.

The above embodiments are merely illustrative of the principles and utilities of the present application and are not intended to limit the present application. Any person skilled in the art can modify or change the above-mentioned embodiments without departing from the spirit and scope of the present application. Accordingly, it is intended that all equivalent modifications or changes which can be made by those skilled in the art without departing from the spirit and technical concepts disclosed in the present application shall be covered by the claims.

Claims (10)

1. A system for policy control, comprising: the system comprises a central control system, at least two robots and a scheduling strategy database; the central control system is provided with a receiving module for receiving input inquiry content; the robot is provided with a communication module for receiving the inquiry content and giving a callback result corresponding to the inquiry content; the central control system is provided with a calling module and a decision module, the calling module is used for calling a preset scheduling strategy in a scheduling strategy database, the decision module is used for making a decision, and the decision comprises selecting a conversation robot meeting the current conversation requirement from the at least two robots and selecting an optimal conversation result from the conversation results of the conversation robots meeting the current conversation requirement; the preset scheduling strategy is configurable; the central control system is provided with a feedback module for feeding back the optimal answer result.

2. The system of claim 1, wherein the scheduling policy database comprises a plurality of predetermined scheduling policies, and the predetermined scheduling policies for invoking the conversation robot comprise personalization information, conversation robot priority, intent, context of a current conversation.

3. The system according to claim 1, wherein the scheduling policy database comprises a plurality of predetermined scheduling policies, and the predetermined scheduling policies for retrieving the optimal callback result comprise a current status of the conversation robot, a confidence score of the callback, and an intention.

4. The system for policy control according to claim 1, wherein said dialogue robot gives a callback result comprising confidence score, current status, intention of the callback result.

5. The system for policy control according to claim 1, wherein the at least two robots are communicatively connected to the central control system for recording real-time status of the at least two robots.

6. The system for policy control according to claim 1 wherein said at least two robots are accessed through a standardized interface.

7. A policy control method using the system according to any one of claims 1 to 6, comprising the steps of:

the receiving module receives input inquiry content;

the central control system calls a preset scheduling strategy, wherein the preset scheduling strategy is a scheduling strategy for selecting the conversation robot;

aiming at input inquiry content, a central control system selects a conversation robot meeting the current conversation requirement from at least two robots according to a scheduling strategy of the selected conversation robot;

sending the input inquiry content to the conversation robot;

the conversation robot receives the inquiry content and gives a callback result;

the central control system calls a preset scheduling strategy, wherein the preset scheduling strategy is a scheduling strategy for selecting an optimal answer result;

aiming at the answer results of the conversation robot, the central control system selects the optimal answer result from the answer results of the conversation robot according to the scheduling strategy for selecting the optimal answer result;

and the central control system feeds back an optimal answer result.

8. The method according to claim 7, wherein the receiving module receives the input query content, the central control system invokes a preset scheduling policy, the preset scheduling policy includes personalized information, a priority of the conversation robot, an intention, and a context of a current conversation, and the central control system makes a decision to select the conversation robot from the at least two robots according to the preset scheduling policy.

9. The policy control method according to claim 7, wherein the conversation robot selects whether to make a call back according to the current status after receiving the inquiry content, and if the call back is made, the communication module of the conversation robot feeds back the result of the call back to the central control system.

10. The strategy control method according to claim 7, characterized in that the central control system receives the answer results of the conversation robot, calls a preset scheduling strategy which comprises the current state of the conversation robot, the confidence score and the intention of the answer, and makes a decision to select the optimal answer result from the answer results of the conversation robot according to the preset scheduling strategy.

Images