A community-scene-oriented multi-service robot digital twin management system

Abstract

The application discloses a community-scene-oriented multi-service robot digital twin management system, belongs to the technical field of digital twins, and is a community-scene-oriented multi-service robot scheduling management system based on digital twins. In order to adapt to the characteristics of different service demand types, dynamic changes in density, demand type diversity, and distribution difference in different time periods in a community scene, the application builds a digital mapping of an actual community scene based on a digital twin prototype platform, provides real-time people flow density information, different type demand density information, and the like, learns and predicts different type service demand density information, and matches scheduling and path planning according to different service robot service characteristics by using a B-spline spline curve and a gradient descent optimization method. The application has the capabilities of multi-service robot pre-scheduling, high demand matching degree scheduling, and optimal scheduling and management in a community scene.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of digital twinning, in particular to a multi-service robot digital twinning management system for community scenarios. BACKGROUND

[0002] Digital twinning is a simulation process integrating multi-disciplinary, multi-physical quantity, multi-scale and multi-probability, which fully utilizes physical models, sensor updates, operation history and other data, and completes mapping in a virtual space, thereby reflecting the whole life cycle process of the corresponding entity equipment. Digital twinning is a concept beyond reality, which can be regarded as a digital mapping system of one or more important and interdependent equipment systems. It has more and more applications in product design, operation management and other aspects today.

[0003] Digital community is a product of the combination of the current digital era and community management, and how to make community management and service more efficient and visible, and reduce operation cost is the research focus of community management. SUMMARY

[0004] The purpose of the present application is to solve the above problems, and provide a multi-service robot digital twinning management system for community scenarios, comprising the following steps:

[0005] S1: building a digital twinning platform;

[0006] S2: completing digital mapping of the actual community scenario based on the digital twinning platform, and providing real-time personnel flow information and constantly updated database service information of the actual community scenario;

[0007] S3: deep learning of the personnel flow information and the database service information, and completing mapping of different characteristics of people to demand types;

[0008] S4: motion planning of the service robot aiming at the highest service type density;

[0009] S5: based on the database service information of the actual community scenario, learning the specific needs of the permanent users, calculating the density distribution of different areas and different types of needs according to the characteristics of the flow, matching the characteristics of different multi-service robots, and completing the pre-scheduling of the multi-service robots in different areas of the actual community in different time periods.

[0010] Further, in S3, the model is updated and learned according to the information in the constantly updated service database by using continuous wavelet transform and deep convolutional neural network, so as to obtain the mapping of the multi-type dimensions of the personnel in the actual community scenario to the demand type distribution.

[0011] Further, in S4, the digital twin platform is used to plan the path of the service robot according to the personnel flow information, and a target function is designed according to the serviceable type of each service robot and the density of the corresponding service type in different areas. An initial motion curve is established based on a B-spline curve model, and the path is optimized to have the maximum service demand density in combination with the target function, so as to ensure that there are more same or corresponding service demands along the path of the service robot.

[0012] Further, in S4, the service types of different service robots are considered, and the demand functions corresponding to each service type are represented as D k (x,y), and the target function of the multi-robot is represented as:

[0013]

[0014] wherein K is the number of robots; S k is the service area of the Kth robot; L k is the path length; λ 1k and λ 2k are the weighting factors of the service demand density coverage and the path length of the Kth robot, respectively;

[0015] The B-spline curve is defined by the following parametric equation:

[0016]

[0017]

[0018] Further, by setting the response area distance d r and the service demand function of each area by different crowds, it is ensured that there are more same or corresponding service demands along the path of the service robot, and the service demand function is:

[0019]

[0020] wherein U att is the attraction degree of the current point to the service robot; q and q people are the current position and the position of the people in the response area, respectively; ε is the response adjustment coefficient; and the attraction degree of each point is the sum of the attraction degrees of all crowds in the response area to the current position, represented as

[0021] Further, a gradient descent target function is designed, and the gradient descent target function is represented as:

[0022]

[0023] Wherein, based on the safety distance s of service robot collision f , define c f = s f -||q-q person ||, finally using BFGS method to solve the planning to further improve the efficiency of service robot, shorten the response time.

[0024] Further, in S5, according to the personnel flow information provided by the digital twin platform, the people flow statistics and people flow feature recognition of the actual community scene are input into the trained model for calculation to obtain the prediction of the demand distribution in the actual community scene, thereby realizing pre-scheduling.

[0025] Further, the prepared multi-service robot is arranged to respond to the dynamic disturbance of the service quantity, and the prepared multi-service robot is dynamically supplemented based on the special scene of the required service surge, ensuring the overall operation efficiency.

[0026] Compared with the prior art, the beneficial effects of the present application are: the present application realizes the mapping of the community scene to the digital world through a multi-service robot digital twin management system for community scenes, and real-time sampling of people flow, time, weather and other information. Through the deep learning method, the character demand feature distribution is imaged, and part of the resident users are accurately described, and then the different service density distribution is predicted according to the current situation, and the multi-service robot is pre-scheduled and uniformly moved according to the highest service type density as the target. Realize efficient, low delay, low energy consumption, can respond to dynamic demand surge of community service ability. BRIEF DESCRIPTION OF DRAWINGS

[0027] Figure 1 It is a kind of multi-service robot digital twin management system structure diagram for community scene. DETAILED DESCRIPTION

[0028] The multi-service robot digital twin management system for community scene will be described in more detail below in conjunction with the schematic diagram, which shows the preferred embodiment of the present application, it should be understood that the person skilled in the art can modify the present application described herein, and still achieve the advantageous effects of the present application, therefore, the following description should be understood as the extensive knowledge of the person skilled in the art, and not as the limitation of the present application.

[0029] In the description of the present application, it should be noted that for orientation words, such as the terms "center", "transverse", "longitudinal", "length", "width", "thickness", "upper", "lower", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", "clockwise", "counterclockwise", etc. indicate the orientation and positional relationship based on the orientation or positional relationship shown in the drawings, which is only for the convenience of describing the present application and simplifying the description, and does not indicate or imply that the devices or elements referred to must have a particular orientation, be constructed and operated in a particular orientation, and cannot be understood as limiting the specific protection scope of the present application.

[0030] As shown in Figure 1 A community scenario-oriented multi-service robot digital twin management system is provided for the services provided and the multi-service robots, which are divided in the category dimension, data and theoretical support is provided through the digital twin system, the mapping of different characteristics of people to demand categories is completed through deep learning of the data collected in daily life, and specific demand learning can also be performed on community resident users. According to the characteristics of the flow, the demand density distribution of different areas and different categories is calculated. The characteristics of different service robots are matched to complete the pre-scheduling of multi-service robots in different areas of the community in different time periods, thereby improving service efficiency and shortening service response time.

[0031] Among them, the mapping of different characteristics of people to demand categories is completed by using continuous wavelet transform (CWT) and deep convolutional neural network (CNN), and the model is updated and learned according to the information in the continuously updated service database. The mapping of demand type distribution in the dimensions of gender, age, time, etc. of the person is obtained. As a kind of general demand category mapping.

[0032] The existing data twin platform is used to plan the path of the multi-service robot. According to the serviceable type of each service robot and the density of the corresponding service type in different areas, a target function is designed, the service types of different service robots are considered, and the demand function corresponding to each service type is represented as D k (x,y), then the target function of the multi-robot integrates the service demand of all robots as:

[0033]

[0034] Where K is the number of robots, S k is the service area of the Kth robot, L k is the path length, and λ 1k and λ 2kare the trade-off factors of service demand density coverage and path length of the Kth robot respectively. The initial motion curve is modeled by B-spline curve, which can be defined by the following parametric equation:

[0035]

[0036]

[0037] The gradient ascent method is used to optimize the path with the largest service demand density combined with the objective function, ensuring that there may be more identical or corresponding service demands along the path. By setting the response area distance d r and the service demand function of each area for different groups of people:

[0038]

[0039] where U att is the attraction degree of the current point to the service robot, q and q people represent the current position and the position of the people in the response area respectively, and ε is the response adjustment coefficient. Thus, the attraction degree of each point is the sum of the attraction degrees of all people in the response area to the current position

[0040] and further design the objective function of gradient descent

[0041]

[0042] where s f is the safety distance of the service robot collision, and c f = s f -||q-q person || is defined, and finally the BFGS method is used to solve the planning to further improve the efficiency of the service robot and shorten the response time.

[0043] Another is to map the precise portrait of the demand distribution of the population according to the service order. In the digital twin system, according to the flow statistics, the flow characteristics are identified, the trained model is input for calculation. The demand distribution in the region is predicted. Thus, pre-scheduling is carried out in advance. In data preparation, it is necessary to collect and organize flow statistics data, including but not limited to GPS data, sensor data, service order data, etc. The data should include timestamp, location, resident population information and service order information. Long short-term memory network (LSTM) is selected and trained to predict the change of service demand over time. The model will output the service demand prediction value of each area in a certain period of time in the future. According to the model prediction result, the corresponding pre-scheduling strategy is executed, including but not limited to: allocating more resources to areas with high predicted demand; for time periods with low demand, reasonably scheduling resources to time periods with high demand; developing service plans in advance for emergency situations, such as coping strategies when sudden events cause demand to surge.

[0044] And the multi-service robot digital twin management system for community scenarios of the present application has a prepared multi-service robot that responds to the dynamic disturbance of the number of services. For example, in the face of natural disasters, a surge in the number of people in the community, and other special situations where the demand for services surges, dynamic replenishment ensures higher overall operating efficiency.

[0045] The above is only a preferred embodiment of the present application, and does not limit the present application in any way. Any person skilled in the art, without departing from the scope of the technical solutions of the present application, can make any form of equivalent replacement or modification of the technical solutions and technical content disclosed by the present application, and still belong to the protection scope of the present application.

Claims

1. A community-scene oriented multi-service robot digital twin management system, characterized in that, Comprise the following steps: S1: build a digital twin platform; S2: based on the digital twin platform, complete the digital mapping of the actual community scene, and provide real-time personnel flow information and constantly updated database service information of the actual community scene; S3: deep learning of personnel flow information and database service information, mapping of different characteristics of people to demand types; S4: the highest target of adapting service type density to make motion planning for service robots; Using the built digital twin platform, according to the personnel flow information, the path planning of the service robot is made, and according to the serviceable type of each service robot and the density of the corresponding service type in different areas, the objective function is designed, the initial motion curve is established with B-spline curve as the model, and the path is optimized with the objective function to ensure that there may be more same or corresponding service demand on the path of the service robot; Considering the different service types of different service robots, the demand function corresponding to each service type is represented as The objective function of the multi-robot will integrate the service demand of all robots as follows: ； wherein, is the number of robots; is the service area of the th robot; is the path length; and are the service demand density coverage and path length trade-off factors for the th robot, respectively; B-spline curve is defined by the following parameter equation: ； ； S5: based on the database service information of the actual community scene, the specific demand learning of the permanent user is made, the density distribution of different areas and different types of demand is calculated according to the characteristics of the flow, the characteristics of different multi-service robots are matched, and the pre-scheduling of multi-service robots in different areas of the actual community in different time periods is completed.

2. The community-scene oriented multi-service robot digital twin management system according to claim 1, wherein, In the S3, by using continuous wavelet transform and deep convolutional neural network, the model is updated and learned according to the information in the constantly updated service database, so as to obtain the mapping of the multi-class dimension of the personnel in the actual community scene to the demand type distribution.

3. The community-scene oriented multi-service robot digital twin management system of claim 1, wherein, By setting the response area distance and the service demand function of each area for different groups of people to ensure that the path of the service robot may have more identical or corresponding service demands, the service demand function is: ； wherein, is the attraction degree of the current point to the service robot; and are the current position and the position of the person in the response area, respectively; is the response adjustment coefficient; the attraction degree of each point is the sum of the attraction degrees of all the people in the response area to the current position, expressed as .

4. The community-scene oriented multi-service robot digital twin management system of claim 3, wherein, The gradient descent objective function is designed, and the gradient descent objective function is represented as: ； Wherein, based on the safety distance of service robot collision , define , finally using BFGS method to solve the completion of planning, to further improve the efficiency of service robots, shorten the response time.

5. The community-scene oriented multi-service robot digital twin management system of claim 2, wherein, In the S5, according to the personnel flow information provided by the digital twin platform, the flow statistics and flow feature recognition of the actual community scene are input into the trained model for calculation to obtain the prediction of the demand distribution in the actual community scene, so as to realize the pre-scheduling in advance.

6. The community-scene oriented multi-service robot digital twin management system according to claim 5, wherein, The pre-prepared multi-service robot is arranged to respond to the dynamic disturbance of the service quantity, and the pre-prepared multi-service robot is based on the dynamic supplement in the special scene of the service surge to ensure the overall operation efficiency.

Images