CN112861377B - Planning identification method and device under condition of observable environmental part - Google Patents
Planning identification method and device under condition of observable environmental part Download PDFInfo
- Publication number
- CN112861377B CN112861377B CN202110264430.4A CN202110264430A CN112861377B CN 112861377 B CN112861377 B CN 112861377B CN 202110264430 A CN202110264430 A CN 202110264430A CN 112861377 B CN112861377 B CN 112861377B
- Authority
- CN
- China
- Prior art keywords
- action
- time
- candidate
- plan
- time constraint
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
- G06Q10/047—Optimisation of routes or paths, e.g. travelling salesman problem
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/04—Constraint-based CAD
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/12—Timing analysis or timing optimisation
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- Human Resources & Organizations (AREA)
- Theoretical Computer Science (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- General Physics & Mathematics (AREA)
- Game Theory and Decision Science (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Development Economics (AREA)
- Computer Hardware Design (AREA)
- Evolutionary Computation (AREA)
- Geometry (AREA)
- General Engineering & Computer Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
One or more embodiments of the present disclosure provide a method and an apparatus for identifying a plan under an environment partially observable condition, including: according to the obtained observation information, determining that the observation information corresponds to candidate actions in a preset planning library; judging whether the candidate action meets a time constraint condition corresponding to a time constraint model or not, and if yes, taking the candidate action as a target action; wherein the time constraint model is determined from the planning library. The planning recognition of the recognized object can be realized, and the recognition result is accurate.
Description
Technical Field
One or more embodiments of the present disclosure relate to the technical field of artificial intelligence, and in particular, to a method and an apparatus for identifying a plan under an environment portion observable condition.
Background
Planning recognition is an important branch of the field of artificial intelligence, and refers to the process of inferring a reasonable and complete planning description for an agent from observations of fragmented, trivial actions or action effects for one or more agents. By planning identification, information which has occurred but has not been observed can be inferred, and future targets or actions which are likely to be executed can also be inferred. The planning identification can be applied to military simulation, multi-agent system cooperation, robot assistance and the like, and has wide application prospect and higher practical value.
Disclosure of Invention
In view of this, one or more embodiments of the present disclosure provide a method and an apparatus for identifying a plan under an environment partially observable condition, which can implement plan identification of an identified object.
In view of the above, one or more embodiments of the present disclosure provide a plan recognition method under an observable environment part condition, including:
according to the obtained observation information, determining that the observation information corresponds to candidate actions in a preset planning library;
judging whether the candidate action meets a time constraint condition corresponding to a time constraint model or not, and if so, taking the candidate action as a target action; wherein the time constraint model is determined from the planning library.
Optionally, the plan library includes at least one layer of plans and at least one action under a sub-plan; determining a time constraint model according to a preset planning library, wherein the time constraint model comprises the following steps:
constructing an action pair consisting of two actions based on all the actions in the planning library; wherein a first action of the pair of actions is performed before a second action;
calculating the time interval of each action pair;
constructing the time constraint model based on the time intervals corresponding to all action pairs and each action pair respectively, the action and the time thereof executed first in time sequence and the action and the time thereof executed last; and each action pair and the corresponding time interval thereof are the time constraint conditions of the action pair, the first executed action and the time thereof are the time constraint conditions of the first executed action, and the last executed action and the time thereof are the time constraint conditions of the last executed action.
Optionally, the calculating a time interval of each action pair includes:
determining a connection path of an action pair according to the execution sequence relation between the first action and the second action;
and calculating the time interval between the first action and the second action according to the connection path of the action pair.
Optionally, the method further includes: encoding the planning library, wherein the encoding method comprises the following steps:
encoding the first layer plan into a single bit number;
adding one bit to the second layer plan on the basis of the first layer plan code, wherein the value of the second bit is increased progressively according to the time sequence relation of each node of the second layer plan;
adding one bit to the N-th layer programming on the basis of the N-1 layer programming code, wherein the value of the (N + 1) -th bit is increased progressively according to the time sequence relation of each node of the N-th layer programming;
for each sub-plan in the nth level plan: for each action under the sub-plan, increasing a digit on the basis of the sub-plan code, and increasing the value of the increased digit of each action according to the time sequence relation.
Optionally, the determining, according to the obtained observation information, that the observation information corresponds to a candidate action in a preset planning library includes:
for the observation information acquired for the first time, searching the planning library according to the observed action to obtain at least one action matched with the planning library, taking the matched at least one action as the candidate action, and adding the code of the candidate action and the current time of the observed action into a candidate action set;
and for the observation information which is not acquired for the first time, searching the planning library according to the observed action to obtain at least one action matched with the planning library, taking the matched at least one action as the candidate action, adding the code of the candidate action and the current time of the observed action into a candidate action set, wherein the candidate action set also comprises the code of the candidate action added last time and the corresponding time.
Optionally, the determining whether the candidate action meets a time constraint condition corresponding to a time constraint model, and if yes, taking the candidate action as a target action includes:
determining a time constraint condition corresponding to the candidate action in the time constraint model for the code of the candidate action added at this time in the candidate action set;
and judging whether the candidate action meets the time constraint condition or not according to the code of the candidate action added at this time and the current time, and if so, taking the candidate action as the target action.
Optionally, determining whether the candidate action meets the time constraint condition according to the code of the candidate action added this time and the current time includes:
and judging whether the candidate action added this time meets the time constraint condition or not according to the code and the current time of the candidate action added this time, the code of the candidate action added last time and the corresponding time for the observation information which is not obtained for the first time.
Optionally, the method further includes:
and constructing a historical action sequence according to the target action under the historical observation.
Optionally, the method further includes:
when the target actions are at least two, sending an inquiry message for inquiring the final target action; and receiving a reply message to the determined final target action.
An embodiment of the present specification further provides a planning identification apparatus under an environment partially observable condition, including:
the candidate action determining module is used for determining that the observation information corresponds to candidate actions in a preset planning library according to the acquired observation information;
the target action determining module is used for judging whether the candidate action meets the time constraint condition corresponding to the time constraint model or not, and if yes, taking the candidate action as a target action; wherein the time constraint model is determined from the planning library.
As can be seen from the above, according to the method and the apparatus for identifying a plan under an observable condition in an environmental part provided in one or more embodiments of the present specification, it is determined that observation information corresponds to a candidate action in a preset plan library according to acquired observation information; and judging whether the candidate action meets the time constraint condition corresponding to the time constraint model, if so, taking the candidate action as a target action, and further determining the planning action executed by the identified object according to the identified target action, thereby realizing the planning identification of the identified object and having more accurate identification result.
Drawings
In order to more clearly illustrate one or more embodiments or prior art solutions of the present specification, the drawings that are needed in the description of the embodiments or prior art will be briefly described below, it is obvious that the drawings in the description below are only one or more embodiments of the present specification, and that other drawings may be obtained by those skilled in the art without inventive effort.
FIG. 1 is a schematic flow chart of a method according to one or more embodiments of the present disclosure;
FIG. 2 is a schematic diagram of a planning library in accordance with one or more embodiments of the present disclosure;
FIG. 3 is a block diagram of an apparatus according to one or more embodiments of the present disclosure;
fig. 4 is a block diagram of an electronic device according to one or more embodiments of the present disclosure.
Detailed Description
For the purpose of promoting a better understanding of the objects, aspects and advantages of the present disclosure, reference is made to the following detailed description taken in conjunction with the accompanying drawings.
It is to be understood that unless otherwise defined, technical or scientific terms used in one or more embodiments of the present disclosure should have the ordinary meaning as understood by one of ordinary skill in the art to which this disclosure belongs. The use of "first," "second," and similar terms in one or more embodiments of the specification is not intended to indicate any order, quantity, or importance, but rather is used to distinguish one element from another. The word "comprising" or "comprises", and the like, means that the element or item listed before the word covers the element or item listed after the word and its equivalents, but does not exclude other elements or items. The terms "connected" or "coupled" and the like are not restricted to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "upper", "lower", "left", "right", and the like are used only to indicate relative positional relationships, and when the absolute position of the object being described is changed, the relative positional relationships may also be changed accordingly.
As shown in fig. 1, one or more embodiments of the present disclosure provide a plan recognition method under an environment portion observable condition, including:
s101: according to the obtained observation information, determining that the observation information corresponds to candidate actions in a preset planning library;
in this embodiment, a planning library is constructed in advance, the planning library describes the decision and planning logic of the identified objects, and each identified object provides a planning library for the observed planning type and the action and sequence of the constituting plan.
The observation information of the identified object can be acquired at a specific time under the condition that the environment is partially observable, and under the condition that the observation information of the identified object at a partial time is lost and cannot be acquired. For example, as shown in table 2, when the method is applied to a military simulation scenario, the observation information of a part of agents may be obtained or the observation information of a part of agents may not be obtained in a battlefield environment.
The observation information is an observed motion of the identified object at the current time. And searching at least one corresponding action in the planning library for the obtained observation information, taking the searched at least one action as a candidate action, and carrying out subsequent judgment on the candidate action.
S102: judging whether the candidate action meets a time constraint condition corresponding to the time constraint model or not, and if so, taking the candidate action as a target action; wherein the time constraint model is determined from a planning library.
In this embodiment, based on the searched candidate action, it is determined whether the candidate action satisfies a time constraint condition corresponding to the candidate action in the time constraint model, and the candidate action satisfying the time constraint condition is used as the identified target action, so that the planning action executed by the identified object can be determined according to the identified target action, thereby implementing the planning identification of the identified object.
In some approaches, based on the planned actions of the identified objects, for a collaborative system scenario (e.g., robot soccer, robot assistance, collaborative combat, etc.), the identified objects may be assisted in performing other actions to complete a collaborative task; for an enemy battle scene, effective actions can be taken in time according to the planning actions of the identified objects to complete the battle mission, which is only an exemplary illustration and is not used to limit the specific application scene and application method.
It is to be appreciated that the method can be performed by any apparatus, device, platform, cluster of devices having computing and processing capabilities.
The plan recognition method of the present specification will be described in detail below with reference to specific embodiments.
In some embodiments, the actions in the planning library should satisfy the time constraint relationship according to the conditions of task execution logic, action time sequence relationship, and the like, and a time constraint model is constructed according to the time constraint relationship between the actions.
In some approaches, the plan library may represent a hierarchical representation of the plan, with the tasks to be performed being decomposed, with nodes of different levels representing the plan of different levels of the identified object, respectively, with the plan represented by each node being decomposed into a combination of plans represented by sub-nodes, with leaf nodes representing actions of the identified object.
The method is applied to military simulation scenes by combining the scheme shown in fig. 2, an action planning library of assault teams is constructed, the first-layer plan (root nodes) is a cross-sea island climbing action plan, the second-layer plan comprises three sub-plans of cross-sea action, landing and attack, the sub-plan cross-sea action comprises three actions of boarding, sailing and launching, the sub-plan landing comprises three actions of maneuvering, observing and cutting wire netting, and the sub-plan attack comprises four actions of maneuvering, observing, aiming and striking.
The action planning library of the assault team has the following time constraint relationship: the sub-plan "sea-crossing maneuver" must be executed before "landing", and the sub-plan "landing" must be executed before "attack". Under the sub-plan "sea-crossing maneuver," go-to-ship "action must be performed before" sail "action, which must be performed before" go-to-ship "action. Under the sub-plan "landing", the "maneuver" action must be performed before the "observe" action and the "cut wire" action, which have no order requirements to be performed. Under the sub-plan "attack," the "aim" action must be performed before the "hit" action, with no order requirements for the other actions to be performed.
And when all actions of the 'sea crossing action', 'landing' and 'attacking' sub-plans in the action planning library are completely finished, the assault tasks of the assault teams are considered to be finished. The friend troops in cooperative operation can identify the current stage and the historical state of the action executed by the assault team through a planning identification method, and can predict the action to be executed in the future, and on the basis, the friend troops in cooperative operation can complete military missions.
In some embodiments, the plan library includes at least one layer of plans and at least one action under a sub-plan; determining a time constraint model according to a preset planning library, wherein the time constraint model comprises the following steps:
constructing an action pair consisting of two actions based on all the actions in the planning library; wherein a first action in the pair of actions is performed before a second action;
calculating the time interval of each action pair;
and constructing a time constraint model based on time intervals corresponding to all action pairs and action pairs, the action executed first in time sequence and the time thereof, and the action executed last in time sequence and the time thereof, wherein each action pair and the corresponding time interval thereof are time constraint conditions of the action pair, the action executed first in time and the time thereof are time constraint conditions of the action executed first in time, and the action executed last in time and the time thereof are time constraint conditions of the action executed last in time.
In this embodiment, the planning library includes a plurality of actions to be executed for completing a specific task, and some actions have a time sequence relationship. Constructing an action pair formed by two actions according to all the actions in the planning library, wherein the first action in the action pair needs to be executed before the second action; after all possible action pairs are constructed, calculating the time interval of each action pair, namely calculating the time interval between the execution of a first action and the execution of a second action, and taking the action pairs and the corresponding time intervals thereof as the time constraint conditions of two actions in the action pairs; based on the time constraint conditions of all action pairs and the time constraint conditions of the action executed first in time sequence (the action executed first and the time thereof), the time constraint conditions of the action executed last (the action executed last and the time thereof) construct a time constraint model, the time constraint model can represent the time constraint relation between the actions, and then constraint judgment is carried out by using the time constraint model according to the observation information of the identified object so as to determine the action executed by the identified object.
In some embodiments, calculating the time interval for each action pair comprises:
determining a connection path of an action pair according to the execution sequence relation between the first action and the second action;
and calculating the time interval between the first action and the second action according to the connection path of the action pair.
The planning identification method of the embodiment provides an assumption condition that the system time is discrete, and the time consumption of each action is 1 basic time, such assumption can be obtained by statistics through expert experience or a machine learning method, and is reasonable in practical application, and the specific process is not described in the specification in a derivation manner.
In some ways, in connection with the embodiment shown in fig. 2, the actions in the plan library include "getting on board", "sailing", "getting off board", "maneuvering", "observing", "cutting wire", "maneuvering", "observing", "aiming", and "striking", and the action pairs that can be constructed include, in a chronological relationship of the actions: (boarding, sailing), (boarding, disembarking), (sailing, disembarking), (boarding, maneuvering), (disembarking, maneuvering), (maneuvering, viewing), (maneuvering, wire netting), (aiming, hitting), etc., the first action of each action pair being performed before the second action.
After discretizing the time of each action, the time for executing each action can be obtained as follows:
TABLE 1 action execution time
| Time of day | Movement of |
| 1 | Boarding ship |
| 2 | Navigation device |
| 3 | Launching boat |
| 4 | Maneuvering |
| 5 | Observation of |
| 6 | Maneuvering |
| 7 | Observation of |
| 8 | Aiming |
| 9 | Shooting device |
The constructed time constraint model supports partially observable conditions of the environment, and observable actions are, for example:
TABLE 2 observable actions
| Time of day | Action of moving |
| 1 | Boarding ship |
| 2 | (node not observable) |
| 3 | Launching ship |
| 4 | Maneuvering |
| 5 | (node not observable) |
| 6 | Maneuvering |
| 7 | (node unobservable) |
| 8 | Aiming |
| 9 | Shooting device |
In some embodiments, calculating a time interval between the first action and the second action according to the connection path of the action pair includes: determining the number of all actions on the connection path, and determining the time interval between the first action and the second action according to the number of all actions. For example, the number of all actions on the connection path is M, and M-1 is taken as the time interval between the first action and the second action.
As shown in fig. 2 and table 1, the connection path of the action pairs (getting on and off) is "getting on, sailing, and off", and there is a node "sailing" between the first action "getting on" and the second action "off", and the time interval between the two actions "getting on" and "off" is 2; the action pairs and the time intervals thereof are expressed as { (go-on-board, go-off-board), 2}, and are added to the time constraint model as the time constraint conditions of the action pairs (go-on-board, go-off-board). For the action pairs (getting on the ship and cutting the wire netting), the connection path is five actions, namely 'getting on the ship, sailing, getting off the ship, maneuvering and cutting the wire netting', a node 'sailing, getting off the ship and maneuvering' exists between the first action 'getting on the ship' and the second action 'cutting the wire netting', and the time interval of the two actions 'getting on the ship' and 'cutting the wire netting' is 4; and (4) representing the action pairs and the time intervals thereof as { (boarding, wire shearing), 4}, and adding the action pairs and the time intervals thereof into the time constraint model as time constraint conditions of the action pairs (boarding and wire shearing).
Based on the planning library shown in fig. 2, it can be determined that the time constraints of each action pair are shown in table 3:
TABLE 3 action pairs and corresponding time intervals
In some embodiments, the plan identification method further comprises: the planning library is encoded. Considering that repeated actions exist in the planning library, the repeated actions can be under the same sub-plan or under different sub-plans, the repeated actions have different execution logics under the same sub-plan or different sub-plans, and the sub-plans and all the actions at different levels have a time sequence relation, and on the basis, the root node, the sub-plans and all the actions are coded based on the planning library.
In some embodiments, encoding the planning library comprises: the first layer of planning codes are a digit, the second layer of planning is added with one digit on the basis of the first layer of planning, namely if the first layer of planning codes corresponding to the root nodes is a digit, each node of the second layer of planning codes is a two digit, each node of the second layer of planning codes increases progressively according to the time sequence relation and the numerical value of the second digit, each layer of planning codes are sequentially planned according to the method, the Nth layer of planning codes is added with one digit on the basis of the (N-1) th layer of planning codes, each node of the Nth layer of planning codes increases progressively according to the time sequence relation and the numerical value of the (N + 1) th digit; if the nth level plan includes at least one sub-plan, for each sub-plan: for each action (leaf node) under the sub-plan, a digit is added on the basis of the sub-plan, and the numerical value of the added digit of each action under the sub-plan is increased according to the time sequence relation.
As shown in fig. 2 and table 3, for example, the root node is coded as 0, and in the second-level plan below the root node, the sub-plan "sea-crossing action" is coded as 00, the sub-plan "landing" is coded as 01, and the sub-plan "attack" is coded as 02; under the sub-plan of the sea-crossing action, the action of going on board is coded as 000, the action of sailing is coded as 001, and the action of going off board is coded as 002; under the sub-plan of 'login', the action 'maneuver' is coded into 010, the action 'observation' is coded into 011, and the action 'wire shearing' is coded into 012; under the sub-program "attack", action "maneuver" is coded as 020, action "observe" is coded as 021, action "aim" is coded as 022, and action "hit" is coded as 023.
By such a coding scheme, a path from the root node to the leaf node can be quickly determined, and the timing of the execution of the action can be identified by the coding, for example, for an action "observation" coded as 011, whose second bit value is 1, an action "aim" coded as 022, whose second bit value is 2, and in a code where the action "observation" and the action "aim", the first bit value is the same, and the second bit value of the action "observation" is smaller than the second bit value of the action "aim", so that it can be determined that the action "observation" is executed no later than the action "aim".
On the basis of a planning library of the coding, a time constraint model of the coding form can be established, and each time constraint condition in the time constraint model of the coding form is represented in the coding form. For example, for the time constraint { (go to ship, shear wire), 4}, it can be expressed as { (000, 012), 4}.
In some embodiments, determining, according to the obtained observation information, that the observation information corresponds to a candidate action in a preset planning library includes:
for the observation information obtained for the first time, searching a planning library according to the observed action to obtain at least one action matched with the planning library, taking the matched at least one action as a candidate action, and adding the code of the candidate action and the current time of the observed action into a candidate action set;
for observation information which is not obtained for the first time, searching a planning library according to observed actions to obtain at least one action matched with the planning library, taking the matched at least one action as a candidate action, adding codes of the candidate actions and the current time of the observed action into a candidate action set, wherein the candidate action set also comprises the codes of the candidate actions added last time and the corresponding time.
In this embodiment, according to the observed ongoing motion of the identified object at the current time, the plan library is searched, the motion matched with the plan library is determined, the searched matched motion is used as a candidate motion, the code of the candidate motion and the current time of the motion are added into the candidate motion set, subsequently, the judgment is performed by using the time constraint model based on the code of each candidate motion in the candidate motion set and the current time of executing the candidate motion, and the target motion meeting the time constraint condition is screened from the candidate motion set.
In some embodiments, determining whether the candidate action meets a time constraint condition corresponding to the time constraint model, and if yes, taking the candidate action as a target action includes:
for each candidate action in the candidate action set, determining a time constraint condition corresponding to the candidate action in the time constraint model;
and judging whether the candidate action meets the time constraint condition or not according to the code of the candidate action and the current time, and if so, taking the candidate action as a target action.
In this embodiment, after a candidate action set is constructed according to observation information, it is sequentially determined whether a candidate action added this time in the candidate action set meets a time constraint condition corresponding to the candidate action in a time constraint model, and a candidate action meeting the time constraint condition is taken as an identified target action. And judging whether the candidate action added this time meets the time constraint condition corresponding to the candidate action added this time or not according to the code and the current time of the candidate action added this time, the candidate action added last time and the time of the candidate action added last time.
With reference to the embodiment shown in fig. 2, in the first observation, the obtained observation information is: taking the motion observed at the moment 1 as 'getting on the ship', searching a planning library according to the motion 'getting on the ship', finding out a candidate motion 'getting on the ship' and a code 000 thereof corresponding to the motion in the planning library, and adding the code 000 and the moment 1 into a candidate motion set; and then, judging whether the time constraint condition related to the 'getting on the ship' action is met or not according to the code 000 and the time 1, wherein the 'getting on the ship' is the first executed action, and judging that the candidate action meets the corresponding time constraint condition according to the time constraint condition of the first executed action, so that the 'getting on the ship' is taken as the identified target action, and further, the identification result of the first observation is that the current plan of the identified object is 'synthetic camping and sea crossing landing action plan-sea crossing maneuvering-getting on the ship'.
In the second observation, the obtained observation information is: taking the motion observed at the moment 3 as 'ship-off', searching a planning library according to the motion 'ship-off', finding a candidate motion 'ship-off' and a code 002 thereof corresponding to the motion in the planning library, and adding the code 002 and the moment 3 into a candidate motion set; and then, judging whether the time constraint conditions of the relevant action pair (getting on the ship and getting off the ship) are met or not according to the code 002, the time 3, the action code 000 and the time (time 1) of the last observation, wherein the time interval between the time 3 and the time 1 is 2, and judging that the candidate action meets the corresponding time constraint condition according to the time constraint conditions { (getting on the ship 000, getting off the ship 002) and 2} of the action pair (getting on the ship and getting off the ship), so that the getting off the ship is taken as the identified target action, and further, the identification result of the second observation is that the current plan of the identified object is 'synthetic operation sea landing action plan-sea crossing maneuver-getting on the ship-sailing-off the ship', thereby realizing the plan identification.
In the third observation, the obtained observation information is: taking the motion observed at the moment 4 as a 'maneuver', searching a planning library according to the motion 'maneuver', finding a candidate motion 'maneuver' and a code 010 thereof, a candidate motion 'maneuver' and a code 020 thereof which correspond to the motion in the planning library, and adding the code 010, the moment 4, the code 020 and the moment 4 into a candidate motion set; at this time, the candidate action set includes the code 002 and the time 3 (previous observation information), the code 010 and the time 4, and the code 020 and the time 4 (current observation information); and then, judging whether the time constraint condition corresponding to the action in the time constraint model is met or not for the code 010 and the moment 4 added at this time, wherein the time constraint condition of the action pair (getting off the ship and maneuvering) is { (002 getting off the ship, 010) and 1}, and judging that the code 010 and the moment 4 of the observation information at this time meet the time constraint condition of the action pair according to the code 002 and the moment 3 of the observation information at the last time and the code 010 and the moment 4 of the observation information at this time, so that the action is taken as the identified target action, and further, the identification result of the third observation is that the current plan of the identified object is 'synthetic camping sea landing plan action-landing-maneuver', thereby realizing the plan identification. And for the code 020 and the time 4 added at this time, because the current observation information and the last observation information do not conform to the time constraint condition { (lower ship 002, maneuver 020), 3} of the action pair (lower ship 002, maneuver), the action is screened out and is not taken as the target action.
In some embodiments, the plan identification method further comprises:
and constructing a historical action sequence according to the target action under the historical observation.
In this embodiment, not only the target motion under the current observation can be identified according to the observation information obtained this time, but also a historical motion sequence formed by the target motions under the historical observation can be constructed according to the target motion under the historical observation, so that the historical planning identification can be realized.
In some embodiments, the plan identification method further comprises:
when the target actions are at least two, sending an inquiry message for inquiring the final target action; and receiving a reply message to the determined final target action.
In this embodiment, when at least two target actions are screened out by using the time constraint model according to the acquired observation information, it is further determined that a final target action is determined from the plurality of target actions, so that the planning of the identified object is determined according to the final target action, and subsequently, the cooperation mode can be determined according to the final target action. In order to determine the final target action, an inquiry message can be sent to the identified object for inquiring the final target action, and after receiving the inquiry message, the identified object sends a reply message including the final target action according to the planning condition of the identified object, so that the final target action is determined according to the reply message. In this way, the plan executed by the identified object can be determined quickly and accurately by means of the query. Furthermore, an inquiry message can be continuously sent to the identified object to inquire whether a specific action needs to be executed or not so as to assist the identified object to complete the planning.
It should be noted that the method of one or more embodiments of the present disclosure may be performed by a single device, such as a computer or server. The method of the embodiment can also be applied to a distributed scene and is completed by the mutual cooperation of a plurality of devices. In such a distributed scenario, one of the devices may perform only one or more steps of the method of one or more embodiments of the present disclosure, and the devices may interact with each other to complete the method.
It should be noted that the above description describes certain embodiments of the present disclosure. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims may be performed in a different order than in the embodiments and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing may also be possible or may be advantageous.
As shown in fig. 3, one or more embodiments of the present disclosure further provide a plan recognition apparatus under an environment partially observable condition, including:
the candidate action determining module is used for determining that the observation information corresponds to candidate actions in a preset planning library according to the obtained observation information;
the target action determining module is used for judging whether the candidate action meets the time constraint condition corresponding to the time constraint model or not, and if yes, taking the candidate action as the target action; wherein the time constraint model is determined from a planning library.
For convenience of description, the above devices are described as being divided into various modules by functions, and are described separately. Of course, the functionality of the various modules may be implemented in the same one or more pieces of software and/or hardware in implementing one or more embodiments of the present description.
The apparatus in the foregoing embodiment is used for implementing the corresponding method in the foregoing embodiment, and has the beneficial effects of the corresponding method embodiment, which are not described herein again.
Fig. 4 is a schematic diagram illustrating a more specific hardware structure of an electronic device according to this embodiment, where the electronic device may include: a processor 1010, a memory 1020, an input/output interface 1030, a communication interface 1040, and a bus 1050. Wherein the processor 1010, memory 1020, input/output interface 1030, and communication interface 1040 are communicatively coupled to each other within the device via bus 1050.
The processor 1010 may be implemented by a general-purpose CPU (Central Processing Unit), a microprocessor, an Application Specific Integrated Circuit (ASIC), or one or more Integrated circuits, and is configured to execute related programs to implement the technical solutions provided in the embodiments of the present disclosure.
The Memory 1020 may be implemented in the form of a ROM (Read Only Memory), a RAM (Random Access Memory), a static Memory device, a dynamic Memory device, or the like. The memory 1020 may store an operating system and other application programs, and when the technical solution provided by the embodiments of the present specification is implemented by software or firmware, the relevant program codes are stored in the memory 1020 and called to be executed by the processor 1010.
The input/output interface 1030 is used for connecting an input/output module to input and output information. The i/o module may be configured as a component within the device (not shown) or may be external to the device to provide corresponding functionality. The input devices may include a keyboard, a mouse, a touch screen, a microphone, various sensors, etc., and the output devices may include a display, a speaker, a vibrator, an indicator light, etc.
The communication interface 1040 is used for connecting a communication module (not shown in the drawings) to implement communication interaction between the present apparatus and other apparatuses. The communication module can realize communication in a wired mode (such as USB, network cable and the like) and also can realize communication in a wireless mode (such as mobile network, WIFI, bluetooth and the like).
It should be noted that although the above-mentioned device only shows the processor 1010, the memory 1020, the input/output interface 1030, the communication interface 1040 and the bus 1050, in a specific implementation, the device may also include other components necessary for normal operation. In addition, those skilled in the art will appreciate that the above-described apparatus may also include only the components necessary to implement the embodiments of the present disclosure, and need not include all of the components shown in the figures.
The electronic device of the foregoing embodiment is used for implementing the corresponding method in the foregoing embodiment, and has the beneficial effects of the corresponding method embodiment, which are not described again here.
Computer-readable media, including both permanent and non-permanent, removable and non-removable media, for storing information may be implemented in any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information that can be accessed by a computing device.
Those of ordinary skill in the art will understand that: the discussion of any embodiment above is meant to be exemplary only, and is not intended to intimate that the scope of the disclosure, including the claims, is limited to these examples; within the spirit of the present disclosure, features from the above embodiments or from different embodiments may also be combined, steps may be implemented in any order, and there are many other variations of different aspects of one or more embodiments of the present description as described above, which are not provided in detail for the sake of brevity.
In addition, well-known power/ground connections to Integrated Circuit (IC) chips and other components may or may not be shown in the provided figures, for simplicity of illustration and discussion, and so as not to obscure one or more embodiments of the disclosure. Furthermore, devices may be shown in block diagram form in order to avoid obscuring the understanding of one or more embodiments of the present description, and this also takes into account the fact that specifics with respect to implementation of such block diagram devices are highly dependent upon the platform within which the one or more embodiments of the present description are to be implemented (i.e., specifics should be well within purview of one skilled in the art). Where specific details (e.g., circuits) are set forth in order to describe example embodiments of the disclosure, it should be apparent to one skilled in the art that one or more embodiments of the disclosure can be practiced without, or with variation of, these specific details. Accordingly, the description is to be regarded as illustrative instead of restrictive.
While the present disclosure has been described in conjunction with specific embodiments thereof, many alternatives, modifications, and variations thereof will be apparent to those skilled in the art in light of the foregoing description. For example, other memory architectures, such as Dynamic RAM (DRAM), may use the discussed embodiments.
It is intended that the one or more embodiments of the present specification embrace all such alternatives, modifications and variations as fall within the broad scope of the appended claims. Therefore, any omissions, modifications, substitutions, improvements, and the like that may be made without departing from the spirit and principles of one or more embodiments of the present disclosure are intended to be included within the scope of the present disclosure.
Claims (8)
1. A plan recognition method under the condition that an environment part is considerable is characterized by comprising the following steps:
according to the obtained observation information, determining that the observation information corresponds to candidate actions in a preset planning library;
judging whether the candidate action meets a time constraint condition corresponding to a time constraint model or not, and if so, taking the candidate action as a target action;
wherein the plan library comprises at least one layer of plans and at least one action under a sub-plan; determining a time constraint model according to a preset planning library, wherein the time constraint model comprises the following steps:
constructing an action pair consisting of two actions based on all the actions in the planning library; wherein a first action of the pair of actions is performed before a second action;
calculating the time interval of each action pair;
constructing the time constraint model based on the time intervals corresponding to all action pairs and each action pair respectively, the action and the time thereof executed at the first time and the action and the time thereof executed at the last time in the time sequence; each action pair and the corresponding time interval thereof are the time constraint conditions of the action pair, the first executed action and the time thereof are the time constraint conditions of the first executed action, and the last executed action and the time thereof are the time constraint conditions of the last executed action;
encoding the planning library, wherein the encoding method comprises the following steps:
encoding the first layer plan into a single bit number;
adding one bit to the second layer plan on the basis of the first layer plan code, wherein the value of the second bit is increased progressively according to the time sequence relation of each node of the second layer plan;
adding one bit to the N-th layer programming on the basis of the N-1 layer programming code, wherein the value of the (N + 1) -th bit is increased progressively according to the time sequence relation of each node of the N-th layer programming;
for each sub-plan in the nth level plan: for each action under the sub-plan, increasing one digit on the basis of the sub-plan code, and increasing the value of the increased digit of each action according to the time sequence relation.
2. The method of claim 1, wherein calculating the time interval for each action pair comprises:
determining a connection path of an action pair according to the execution sequence relation between the first action and the second action;
and calculating the time interval between the first action and the second action according to the connection path of the action pair.
3. The method of claim 1, wherein the determining that the observation information corresponds to a candidate action in a preset planning library according to the obtained observation information comprises:
for the observation information acquired for the first time, searching the planning library according to the observed action to obtain at least one action matched with the planning library, taking the matched at least one action as the candidate action, and adding the code of the candidate action and the current time of the observed action into a candidate action set;
and for the observation information which is not acquired for the first time, searching the planning library according to the observed action to obtain at least one action matched with the planning library, taking the matched at least one action as the candidate action, adding the code of the candidate action and the current time of the observed action into a candidate action set, wherein the candidate action set also comprises the code of the candidate action added last time and the corresponding time.
4. The method of claim 3, wherein the determining whether the candidate action meets a time constraint condition corresponding to a time constraint model, and if yes, taking the candidate action as a target action comprises:
determining a time constraint condition corresponding to the candidate action in the time constraint model for the code of the candidate action added at this time in the candidate action set;
and judging whether the candidate action meets the time constraint condition or not according to the code of the candidate action added at this time and the current time, and if so, taking the candidate action as the target action.
5. The method of claim 4, wherein determining whether the candidate action meets the time constraint condition according to the coding of the candidate action added this time and the current time comprises:
and judging whether the candidate action added this time meets the time constraint condition or not according to the code and the current time of the candidate action added this time, the code of the candidate action added last time and the corresponding time for the observation information which is not obtained for the first time.
6. The method of claim 1, further comprising:
and constructing a historical action sequence according to the target action under the historical observation.
7. The method of claim 1, further comprising:
when the target actions are at least two, sending an inquiry message for inquiring a final target action; and receiving a reply message to the determined final target action.
8. A plan recognition apparatus for conditions where an environmental portion is observable, comprising:
the candidate action determining module is used for determining that the observation information corresponds to candidate actions in a preset planning library according to the obtained observation information;
the target action determining module is used for judging whether the candidate action meets the time constraint condition corresponding to the time constraint model or not, and if yes, taking the candidate action as a target action; wherein the plan library comprises at least one layer of plans and at least one action under a sub-plan; determining a time constraint model according to a preset planning library, wherein the time constraint model comprises the following steps:
constructing an action pair consisting of two actions based on all the actions in the planning library; wherein a first action of the pair of actions is performed before a second action;
calculating the time interval of each action pair;
constructing the time constraint model based on the time intervals corresponding to all action pairs and each action pair respectively, the action and the time thereof executed first in time sequence and the action and the time thereof executed last; each action pair and the corresponding time interval thereof are the time constraint conditions of the action pair, the first executed action and the time thereof are the time constraint conditions of the first executed action, and the last executed action and the time thereof are the time constraint conditions of the last executed action;
encoding the planning library, wherein the encoding method comprises the following steps:
encoding the first layer plan into a single bit number;
adding one bit to the second layer plan on the basis of the first layer plan code, wherein the value of the second bit is increased progressively according to the time sequence relation of each node of the second layer plan;
adding one bit to the N layer plan on the basis of the N-1 layer plan code, wherein the value of the (N + 1) th bit is increased progressively according to the time sequence relation of each node of the N layer plan;
for each sub-plan in the N-th level plan: for each action under the sub-plan, increasing a digit on the basis of the sub-plan code, and increasing the value of the increased digit of each action according to the time sequence relation.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110264430.4A CN112861377B (en) | 2021-03-11 | 2021-03-11 | Planning identification method and device under condition of observable environmental part |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110264430.4A CN112861377B (en) | 2021-03-11 | 2021-03-11 | Planning identification method and device under condition of observable environmental part |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112861377A CN112861377A (en) | 2021-05-28 |
| CN112861377B true CN112861377B (en) | 2023-02-10 |
Family
ID=75994001
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110264430.4A Active CN112861377B (en) | 2021-03-11 | 2021-03-11 | Planning identification method and device under condition of observable environmental part |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112861377B (en) |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109634304A (en) * | 2018-12-13 | 2019-04-16 | 中国科学院自动化研究所南京人工智能芯片创新研究院 | Unmanned plane during flying paths planning method, device and storage medium |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070226164A1 (en) * | 2006-03-21 | 2007-09-27 | Honeywell International Inc. | Type variables and/or temporal constraints in plan recognition |
| CN101494535A (en) * | 2009-03-05 | 2009-07-29 | 范九伦 | Method for constructing network inbreak scene based on hidden Mrakov model |
| CN102867224A (en) * | 2012-09-03 | 2013-01-09 | 中国科学院自动化研究所 | Method for predicating expected effectiveness of behavior plan |
| JP2019212245A (en) * | 2018-06-08 | 2019-12-12 | 富士通株式会社 | Simulation program, simulation method and simulation device |
| CN109657868B (en) * | 2018-12-26 | 2021-02-09 | 北京理工大学 | Probability planning and identifying method for task time sequence logic constraint |
-
2021
- 2021-03-11 CN CN202110264430.4A patent/CN112861377B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109634304A (en) * | 2018-12-13 | 2019-04-16 | 中国科学院自动化研究所南京人工智能芯片创新研究院 | Unmanned plane during flying paths planning method, device and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112861377A (en) | 2021-05-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111507150B (en) | Method for identifying human face by utilizing multiple image block combination based on deep neural network | |
| CN108645411B (en) | Robot path planning method and device based on particle swarm algorithm and terminal equipment | |
| KR20210076886A (en) | Parking model generation method and device, electronic device and storage medium | |
| Jalali et al. | Autonomous robot navigation system using the evolutionary multi-verse optimizer algorithm | |
| CN114047764B (en) | Training method of path planning model, path planning method and device | |
| CN112561395A (en) | Unmanned aerial vehicle cooperation method, system, device, electronic equipment and storage medium | |
| CN111982138B (en) | Prediction model obtaining and path planning method, device and storage medium | |
| Luo et al. | Jfp: Joint future prediction with interactive multi-agent modeling for autonomous driving | |
| CN107146061A (en) | A kind of ship Guakao event method for digging and system based on harbour multi-layer area | |
| Parhi et al. | Navigational strategy for underwater mobile robot based on adaptive neuro-fuzzy inference system model embedded with shuffled frog leaping algorithm–based hybrid learning approach | |
| CN112861377B (en) | Planning identification method and device under condition of observable environmental part | |
| CN110749325B (en) | Flight path planning method and device | |
| CN109003329A (en) | A kind of target goods heap monitoring device and storage medium | |
| CN116258252A (en) | Aircraft track prediction method, storage medium and equipment based on attention mechanism | |
| Xu et al. | Context-aware timewise vaes for real-time vehicle trajectory prediction | |
| Chen et al. | Q‐EANet: Implicit social modeling for trajectory prediction via experience‐anchored queries | |
| CN113743583A (en) | Intelligent agent invalid behavior switching inhibition method based on reinforcement learning | |
| CN112989610A (en) | Model-driven simulation verification method | |
| US20200372080A1 (en) | Real-time tree search with pessimistic survivability trees | |
| CN114020041B (en) | Multi-unmanned aerial vehicle multi-thread two-dimensional exploration simulation method and system | |
| Huang et al. | Multi-object Detection, Tracking and Prediction in Rugged Dynamic Environments | |
| Kruglikov | A problem of optimal control and observation for distributed homogeneous multi-agent system | |
| CN113642448B (en) | Image recognition method and device for sea/ground maneuvering target by air platform | |
| CN115487508B (en) | Training method and related device for game team recommendation model | |
| CN111817770B (en) | Relay unmanned aerial vehicle deployment method and terminal equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |