A resource scheduling method and system based on deep learning for intelligent property management

Abstract

The present invention relates to the field of deep learning technology, and discloses a method and system for implementing resource scheduling in smart properties based on deep learning. The method comprises: using owner housing information to calculate the owner maintenance probability of the owner, using tenant housing information to calculate the tenant maintenance probability of the tenant, and using the owner maintenance probability and the tenant maintenance probability to schedule human resources, thereby obtaining a human resource scheduling result for the human resources; using equipment usage information to calculate the equipment update probability of public equipment, and using the equipment update probability to schedule equipment resources, thereby obtaining an equipment scheduling result for the equipment resources; determining the data scheduling actions of cell users on data resources, using action influencing factors to calculate the data scheduling probability of the data resources, and using the data scheduling probability to schedule the data resources, thereby obtaining a data scheduling result for the data resources. The present invention can consider complex cell data for resource scheduling.

Description

Technical Field

[0001] The present invention relates to the field of deep learning technology, and in particular to a method and system for implementing resource scheduling in smart properties based on deep learning. Background Art

[0002] Realizing resource scheduling under smart properties based on deep learning refers to the process of using deep learning technology to predict and analyze the schedulable resources under smart properties, and scheduling property resources for community users based on the prediction and analysis results.

[0003] Currently, using deep learning technology to predict and analyze dispatchable resources in smart property management can be divided into two aspects. One is to use deep learning data prediction from the perspective of property resources, analyzing the scheduling trends of property resources in future periods, thereby rationally allocating property resources. The other is to analyze the time-varying trends of community users' use of certain resources, thereby allocating resources to community users when they need them. It can be seen that both methods analyze linear trends over time. However, in reality, community environments are complex, and the reasons for dispatching a certain property resource are also ever-changing. In reality, not all data can be predicted using trend charts. Therefore, a resource scheduling method that can consider complex community data is urgently needed. Summary of the Invention

[0004] In order to solve the above problems, the present invention provides a resource scheduling method and system for smart properties based on deep learning, which can consider complex cell data for resource scheduling.

[0005] In a first aspect, the present invention provides a method for resource scheduling in smart properties based on deep learning, comprising:

[0006] Collect human resources, equipment resources, and data resources under the smart property, and divide the community users under the smart property into owners and tenants;

[0007] collecting the owner's housing information of the owner and the tenant's housing information of the tenant, calculating the owner's maintenance probability of the owner using the owner's housing information, calculating the tenant's maintenance probability of the tenant using the tenant's housing information, and scheduling the human resources using the owner's maintenance probability and the tenant's maintenance probability to obtain a human resource scheduling result of the human resources;

[0008] collecting public device videos of users in the cell, identifying device usage information of the public devices in the public device videos, calculating a device update probability of the public devices using the device usage information, and scheduling the device resources using the device update probability to obtain a device scheduling result for the device resources;

[0009] Determine the data scheduling actions of the users in the cell for the data resources, where the data scheduling actions include modification, addition, deletion, and query. Collect the action influencing factors of the data scheduling actions, calculate the data scheduling probability of the data resources using the action influencing factors, and schedule the data resources using the data scheduling probability to obtain the data scheduling result of the data resources;

[0010] Take the manpower scheduling result, the equipment scheduling result, and the data scheduling result as the resource scheduling result under the intelligent property management.

[0011] In a possible implementation manner of the first aspect, the collection of the human resources, equipment resources, and data resources under the intelligent property management includes:

[0012] Collect the human data, equipment data, and data data under the intelligent property management;

[0013] Based on the human data, calculate the human resources using the following formula:

[0014]

[0015] where, represents the human resources belonging to the i-th property area, m represents the total number of property human resources in the human data that belong to the i-th property area and are in the idle state rest, represents the j-th property human resource in the human data that belongs to the i-th property area and is in the idle state rest;

[0016] Based on the equipment data, calculate the equipment resources using the following formula:

[0017]

[0018] where, represents the equipment resources of the k-th type, S represents the total number of devices of the k-th type in the equipment data, represents the s-th device of the k-th type in the equipment data;

[0019] Based on the data data, calculate the data resources using the following formula:

[0020]

[0021] where, represents the data resources belonging to the i-th property area, L represents the total number of community users in the data data that belong to the i-th property area, represents the data quantity of the l-th community user in the data data that belongs to the i-th property area.

[0022] In a possible implementation of the first aspect, calculating the owner maintenance probability of the owner side by using the owner housing information includes:

[0023] Distinguish the hard decoration housing information and the soft decoration housing information in the owner housing information;

[0024] Calculate the hard decoration maintenance probability of the hard decoration housing information by using the following formula:

[0025] p(a1,a2,a3) = P(a1)P(a2∣a1)P(a3∣a1,a2)

[0026]

[0027] Among them, p(b1|a1,a2,a3) represents the hard decoration maintenance probability, a1 represents the housing construction duration in the hard decoration housing information, a2 represents the housing damage degree in the hard decoration housing information, a3 represents the owner category in the hard decoration housing information, p(a1,a2,a3) represents the probability that a1, a2, and a3 occur simultaneously, P(a1) represents the probability of a1 occurring, P(a2∣a1) represents the conditional probability of a2 occurring on the premise that a1 occurs, P(a3∣a1,a2) represents the conditional probability of a3 occurring on the premise that a1 and a2 occur, b1 represents the event of hard decoration maintenance occurring, and p((a1,a2,a3)∩b1) represents the probability that a1, a2, a3 and b1 occur simultaneously;

[0028] Calculate the soft decoration maintenance probability of the soft decoration housing information by using the following formula:

[0029] p(a′1,a3) = P(a′1)P(a3∣a′1)

[0030]

[0031] Among them, p(b2|a′1,a3) represents the soft decoration maintenance probability, a′1 represents the housing decoration duration in the soft decoration housing information, a3 represents the owner category in the soft decoration housing information, p(a′1,a3) represents the probability that a′1 and a3 occur simultaneously, P(a′1) represents the probability of a′1 occurring, P(a3∣a′1) represents the conditional probability of a3 occurring on the premise that a′1 occurs, b2 represents the event of soft decoration maintenance occurring, and p((a′1,a3)∩b2) represents the probability that a′1, a3 and b2 occur simultaneously;

[0032] Based on the hard decoration maintenance probability and the soft decoration maintenance probability, calculate the owner maintenance probability of the owner side.

[0033] In a possible implementation of the first aspect, calculating the homeowner maintenance probability of the homeowner party based on the hard decoration maintenance probability and the soft decoration maintenance probability includes:

[0034] Based on the hard decoration maintenance probability and the soft decoration maintenance probability, calculate the homeowner maintenance probability of the homeowner party using the following formula:

[0035] P(b1) = p(b1|a1,a2,a3)

[0036] P(b2) = p(b2|a′1,a3)

[0037]

[0038] Among them, p(b3|b1,b2) represents the homeowner maintenance probability, p(b1|a1,a2,a3) represents the hard decoration maintenance probability, p(b2|a′1,a3) represents the soft decoration maintenance probability, p(b1,b2,b3) represents the probability that b1, b2, and b3 occur simultaneously, p(b1,b2) represents the probability that b1 and b2 occur simultaneously, and P(b2∣b1) represents the conditional probability that b2 occurs on the premise that b1 occurs.

[0039] In a possible implementation of the first aspect, calculating the tenant maintenance probability of the tenant party using the tenant housing information includes:

[0040] Calculate the hard decoration maintenance probability corresponding to the tenant housing information using the following formula:

[0041] p(a1,a2,a4) = P(a1)P(a2∣a1)P(a4∣a1,a2)

[0042]

[0043] Among them, p(b4|a1,a2,a4) represents the hard decoration maintenance probability corresponding to the tenant housing information, a1 represents the housing construction duration in the hard decoration housing information, a2 represents the housing damage degree in the hard decoration housing information, a4 represents the tenant category in the tenant housing information, p(a1,a2,a4) represents the probability that a1, a2, and a4 occur simultaneously, P(a1) represents the probability that a1 occurs, P(a2∣a1) represents the conditional probability that a2 occurs on the premise that a1 occurs, P(a4∣a1,a2) represents the conditional probability that a4 occurs on the premise that a1 and a2 occur, b4 represents the event of hard decoration maintenance corresponding to the tenant housing information, and p((a1,a2,a4)∩b4) represents the probability that a1, a2, a4, and b4 occur simultaneously;

[0044] Calculate the soft decoration maintenance probability corresponding to the tenant housing information using the following formula:

[0045] p(a′1,a4)=P(a′1)P(a4|a′1)

[0046]

[0047] Wherein, p(b5|a′1,a3) represents the probability of soft decoration maintenance corresponding to the tenant housing information, a′1 represents the duration of residential decoration in the soft decoration housing information, a4 represents the tenant category in the tenant housing information, p(a′1,a4) represents the probability of a′1 and a4 occurring simultaneously, P(a′1) represents the probability of a′1 occurring, P(a4|a′1) represents the conditional probability of a4 occurring under the premise of a′1 occurring, b5 represents the event of soft decoration maintenance, and p((a′1,a3)∩b5) represents the probability of a′1, a4 and b5 occurring simultaneously;

[0048] The tenant maintenance probability of the tenant is calculated based on the hard decoration maintenance probability corresponding to the tenant housing information and the soft decoration maintenance probability corresponding to the tenant housing information.

[0049] In a possible implementation of the first aspect, identifying device usage information of a public device in the public device video includes:

[0050] Identifying, from the public device video, a usage time of the public device when the device is not updated;

[0051] Calculate the equipment damage probability of the public equipment using a preset YOLOv5s network model;

[0052] Identifying the degree of damage to the public equipment using the equipment damage probability;

[0053] The device usage time and the device damage degree are used as the device usage information.

[0054] In a possible implementation of the first aspect, calculating the device update probability of the public device by using the device usage information includes:

[0055] Obtaining a public device video corresponding to the device usage information;

[0056] The user-device interaction feature between the cell user and the public device is extracted from the public device video using the following formula:

[0057]

[0058] Where B represents the user-device interaction feature, I i(x, y) represents the i-th video frame in the public device video, where i~n represent the sequence numbers of the first n video frames among the first n + 1 video frames, α represents the binary threshold for separating the foreground region and the background region in the (n + 1)-th video frame, and I n+1 (x, y) represents the (n + 1)-th video frame in the public device video, A n+1 (x, y) represents the foreground region in the (n + 1)-th video frame, (A n+1 (x, y)) x represents the gradient of the foreground region in the (n + 1)-th video frame in the horizontal direction, (A n+1 (x, y)) y represents the gradient of the foreground region in the (n + 1)-th video frame in the vertical direction, and G(x, y) represents the gradient magnitude of the foreground region in the (n + 1)-th video frame. represents the ordinate value of the center point of the foreground region in the (n + 1)-th video frame. represents the minimum value of the ordinate values of the center points of the foreground regions in the i~n + 1 video frames. represents the ordinate value of the center point of the foreground region in the i-th video frame;

[0059] Obtain the device usage duration and the device damage degree in the device usage information;

[0060] Based on the device usage duration, the device damage degree, and the user-device interaction characteristics, use the following formula to calculate the device update probability of the public device:

[0061] f = sign(w T φ + b)

[0062] where f represents the device update probability, w represents the weight, b represents the bias, and φ represents a linear function composed of the device usage duration, the device damage degree, and the user-device interaction characteristics.

[0063] In a possible implementation manner of the first aspect, the using the device update probability to perform resource scheduling on the device resources to obtain the device scheduling result of the device resources includes:

[0064] Use the device update probability to determine whether a device anomaly occurs in the public device;

[0065] When a device anomaly occurs in the public device, use the following formula to calculate the device reliability of the public device:

[0066]

[0067] Among them, Reliability represents the device reliability, β and γ represent constant parameters calculated based on the conditional probability when device anomalies occur in the public device, and t represents the time variable;

[0068] Based on the device reliability, use the following formula to calculate the device update time of the device resources:

[0069] Reliability≥δ

[0070]

[0071] Among them, t max represents the device update time, δ represents the threshold of the device reliability, Reliability represents the device reliability, β and γ represent constant parameters calculated based on the conditional probability when device anomalies occur in the public device, and t represents the time variable;

[0072] Perform resource scheduling on the device resources within the device update time to obtain the device scheduling result of the device resources.

[0073] In a possible implementation manner of the first aspect, the performing resource scheduling on the device resources within the device update time to obtain the device scheduling result of the device resources includes:

[0074] Query whether the device resources are in an idle state within the device update time;

[0075] When the device resources are in an idle state within the device update time, identify the current position of the device resources;

[0076] Determine whether the device resources are contended by other preset public devices;

[0077] When the device resources are contended by other preset public devices, calculate the property-target distance and property-other distance between the property area under the intelligent property management and the target public device and the other public device respectively;

[0078] When the sum of the property-target distance and the property-other distance is the minimum value, store the device resources in the property area to perform resource scheduling on the device resources to obtain the device scheduling result of the device resources;

[0079] When the device resources are not contended by other preset public devices, calculate the property-current distance and property-public distance between the property area under the intelligent property management and the current position and the target public device respectively;

[0080] When the sum of the property-current distance and the property-public distance is a minimum, storing the device resources in the property area to perform resource scheduling on the device resources and obtain a device scheduling result for the device resources;

[0081] When the device resource is not in an idle state within the device update time, after delaying the device update time, the process returns to the above step of querying whether the device resource is in an idle state within the device update time.

[0082] In a second aspect, the present invention provides a resource scheduling system for smart properties based on deep learning, the system comprising:

[0083] A user segmentation module is used to collect human resources, equipment resources, and data resources under the smart property, and to divide the users of the community under the smart property into owners and tenants;

[0084] a manpower scheduling module, configured to collect the owner's housing information of the owner and the tenant's housing information of the tenant, calculate the owner's maintenance probability of the owner using the owner's housing information, calculate the tenant's maintenance probability of the tenant using the tenant's housing information, and schedule the human resources using the owner's maintenance probability and the tenant's maintenance probability to obtain a manpower scheduling result for the human resources;

[0085] a device scheduling module, configured to collect videos of public devices of users in the cell, identify device usage information of the public devices in the videos, calculate a device update probability of the public devices using the device usage information, schedule the device resources using the device update probability, and obtain a device scheduling result for the device resources;

[0086] a data scheduling module, configured to determine a data scheduling action of the cell user on the data resource, wherein the data scheduling action includes modification, addition, deletion, and query; collect action influencing factors of the data scheduling action; calculate a data scheduling probability of the data resource using the action influencing factors; schedule the data resource using the data scheduling probability, and obtain a data scheduling result for the data resource;

[0087] The result determination module is used to use the manpower scheduling result, the equipment scheduling result and the data scheduling result as the resource scheduling result under the smart property.

[0088] Compared with the existing technology, the technical principle and beneficial effects of this solution are:

[0089] Embodiments of the present invention collect human resources, equipment resources, and data resources under intelligent property management for summarizing the schedulable resources under the intelligent property management. Further, embodiments of the present invention calculate the owner maintenance probability of the owner side by using the owner housing information to predict the probability that the residence needs to be repaired by human resources under the existing housing information of the owner through conditional probability. Embodiments of the present invention identify the equipment usage information of public equipment in the public equipment video to detect whether the equipment is damaged by using deep learning. Further, embodiments of the present invention calculate the equipment update probability of the public equipment by using the equipment usage information to identify abnormal human body posture features from the public equipment video. Further, embodiments of the present invention perform resource scheduling on the equipment resources by using the equipment update probability to replace the equipment in time when equipment anomalies are detected through the equipment update probability. Therefore, a resource scheduling method and system under intelligent property management based on deep learning proposed by embodiments of the present invention can consider complex community data for resource scheduling. BRIEF DESCRIPTION OF THE DRAWINGS

[0090] The drawings herein are incorporated into the specification and form a part of the specification, showing embodiments consistent with the present invention and used together with the specification to explain the principles of the present invention.

[0091] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, for those of ordinary skill in the art, other drawings can also be obtained based on these drawings without creative efforts.

[0092] Figure 1 It is a flowchart of a resource scheduling method under intelligent property management based on deep learning provided by an embodiment of the present invention;

[0093] Figure 2 It is a module diagram of a resource scheduling system under intelligent property management based on deep learning provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0094] It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0095] An embodiment of the present invention provides a resource scheduling method for intelligent property based on deep learning. The execution subject of the resource scheduling method for intelligent property based on deep learning includes, but is not limited to, at least one of electronic devices such as a server, a terminal, etc. that can be configured to execute the method provided by the embodiment of the present invention. In other words, the resource scheduling method for intelligent property based on deep learning can be executed by software or hardware installed on a terminal device or a server device, and the software can be a blockchain platform. The server includes, but is not limited to: a single server, a server cluster, a cloud server, or a cloud server cluster, etc. The server can be an independent server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, Content Delivery Network (CDN), and big data and artificial intelligence platforms.

[0096] Refer to Figure 1 As shown, it is a schematic flowchart of a resource scheduling method for intelligent property based on deep learning provided by an embodiment of the present invention. Among them, Figure 1 The resource scheduling method for intelligent property based on deep learning described in

[0097] S1. Collect human resources, equipment resources, and material resources under intelligent property, and divide the community users under intelligent property into the owner side and the tenant side.

[0098] In the embodiment of the present invention, by collecting human resources, equipment resources, and material resources under intelligent property, it is used to summarize the schedulable resources under intelligent property.

[0099] Among them, the human resources refer to maintenance personnel for repairing housing walls, water and electricity facilities, and furniture. The equipment resources refer to equipment such as children's play facilities, elderly rehabilitation facilities, trash cans, and public seats in the community. The material resources include medical materials, housing registration information, and registration for leisure activities. It should be noted that the human resources, equipment resources, and material resources are all spare emergency resources, that is, the human resources, equipment resources, and material resources have not been put into use.

[0100] In an embodiment of the present invention, the collection of human resources, equipment resources, and material resources under intelligent property includes: collecting human data, equipment data, and material data under intelligent property; based on the human data, calculating the human resources using the following formula:

[0101]

[0102] Among them, denote the human resources belonging to the i-th property area, and m denotes the total number of property human resources in the human data that belong to the i-th property area and are in the idle state of rest. denote the j-th property human resource in the human data that belongs to the i-th property area and is in the idle state of rest;

[0103] Based on the equipment data, calculate the equipment resources using the following formula:

[0104]

[0105] where, denote the equipment resources of the k-th type, S denotes the total number of equipment of the k-th type in the equipment data, denote the s-th equipment of the k-th type in the equipment data;

[0106] Based on the material data, calculate the material resources using the following formula:

[0107]

[0108] where, denote the material resources belonging to the i-th property area, L denotes the total number of community users in the material data that belong to the i-th property area, denote the material quantity of the l-th community user in the material data that belongs to the i-th property area.

[0109] Among them, the property areas include repairing housing walls, repairing water and electricity facilities, repairing furniture, medical fields, housing registration information, leisure activity registration, etc.

[0110] S2. Collect the housing information of the property owners on the owner side and the housing information of the tenants on the tenant side, calculate the owner maintenance probability of the owner side using the owner housing information, calculate the tenant maintenance probability of the tenant side using the tenant housing information, and schedule the human resources using the owner maintenance probability and the tenant maintenance probability to obtain the human resource scheduling result.

[0111] In the embodiments of the present invention, the owner housing information includes the duration from the date when the house was completed to the present, the duration from the date when the decoration was completed to the present, the population composition of the housing users, the damaged area of the residence, and the number of damaged residences. The tenant housing information includes the duration from the date when the house was completed to the present, the duration from the date when the decoration was completed to the present, the population composition of the housing users, the damaged area of the residence, and the number of damaged residences.

[0112] Furthermore, the embodiment of the present invention calculates the owner's maintenance probability by using the owner's housing information, so as to predict the probability that the residence requires maintenance by human resources under the owner's existing housing information through conditional probability.

[0113] The owner maintenance probability refers to the probability that the house needs to be repaired by human resources based on the owner's existing housing information.

[0114] In one embodiment of the present invention, the calculating of the owner's maintenance probability using the owner's housing information includes: distinguishing between hard furnishings and soft furnishings in the owner's housing information; and calculating the hard furnishing maintenance probability of the hard furnishings information using the following formula:

[0115] p(a1,a2,a3)=P(a1)P(a2∣a1)P(a3∣a1,a2)

[0116]

[0117] Where p(b1|a1,a2,a3) represents the probability of hard decoration maintenance, a1 represents the duration of residential construction in the hard decoration housing information, a2 represents the degree of residential damage in the hard decoration housing information, a3 represents the owner category in the hard decoration housing information, p(a1,a2,a3) represents the probability of a1, a2, and a3 occurring simultaneously, P(a1) represents the probability of a1 occurring, P(a2|a1) represents the conditional probability of a2 occurring under the premise of a1 occurring, P(a3|a1,a2) represents the conditional probability of a3 occurring under the premise of a1 and a2 occurring, b1 represents the event of hard decoration maintenance, and p((a1,a2,a3)∩b1) represents the probability of a1, a2, a3 and b1 occurring simultaneously.

[0118] The soft decoration maintenance probability of the soft decoration housing information is calculated using the following formula:

[0119] p(a′1,a3)=P(a′1)P(a3|a′1)

[0120]

[0121] Where p(b2|a′1,a3) represents the probability of soft decoration maintenance, a′1 represents the duration of residential decoration in the soft decoration housing information, a3 represents the owner category in the soft decoration housing information, p(a′1,a3) represents the probability of a′1 and a3 occurring simultaneously, P(a′1) represents the probability of a′1 occurring, P(a3|a′1) represents the conditional probability of a3 occurring under the premise of a′1 occurring, b2 represents the event of soft decoration maintenance, and p((a′1,a3)∩b2) represents the probability of a′1, a3 and b2 occurring simultaneously;

[0122] The owner maintenance probability of the owner is calculated based on the hard decoration maintenance probability and the soft decoration maintenance probability.

[0123] In another embodiment of the present invention, the calculating the owner maintenance probability of the owner side based on the hard decoration maintenance probability and the soft decoration maintenance probability includes: calculating the owner maintenance probability of the owner side based on the hard decoration maintenance probability and the soft decoration maintenance probability using the following formula:

[0124] P(b1)=p(b1|a1,a2,a3)

[0125] P(b2)=p(b2|a′1,a3)

[0126]

[0127] Among them, p(b3|b1,b2) represents the probability of owner maintenance, p(b1|a1,a2,a3) represents the probability of hard decoration maintenance, p(b2|a′1,a3) represents the probability of soft decoration maintenance, p(b1,b2,b3) represents the probability of b1, b2, and b3 occurring simultaneously, p(b1,b2) represents the probability of b1 and b2 occurring simultaneously, and P(b2|b1) represents the conditional probability of b2 occurring given the occurrence of b1.

[0128] In one embodiment of the present invention, the calculating of the tenant maintenance probability of the tenant by using the tenant housing information includes: calculating the hardware maintenance probability corresponding to the tenant housing information by using the following formula:

[0129] p(a1,a2,a4)=P(a1)P(a2∣a1)P(a4∣a1,a2)

[0130]

[0131] Among them, p(b4|a1,a2,a4) represents the probability of hard decoration maintenance corresponding to the tenant housing information, a1 represents the residential construction duration in the hard decoration housing information, a2 represents the degree of damage to the residential building in the hard decoration housing information, a4 represents the tenant category in the tenant housing information, p(a1,a2,a4) represents the probability of a1, a2, and a4 occurring simultaneously, P(a1) represents the probability of a1 occurring, P(a2|a1) represents the conditional probability of a2 occurring under the premise of a1 occurring, P(a4|a1,a2) represents the conditional probability of a4 occurring under the premise of a1 and a2 occurring, b4 represents the event of hard decoration maintenance corresponding to the tenant housing information, and p((a1,a2,a4)∩b4) represents the probability of a1, a2, a4 and b4 occurring simultaneously;

[0132] The following formula is used to calculate the probability of soft furnishing maintenance corresponding to the tenant's housing information:

[0133] p(a′1,a4)=P(a′1)P(a4|a′1)

[0134]

[0135] Wherein, p(b5|a′1,a3) represents the probability of soft decoration maintenance corresponding to the tenant housing information, a′1 represents the duration of residential decoration in the soft decoration housing information, a4 represents the tenant category in the tenant housing information, p(a′1,a4) represents the probability of a′1 and a4 occurring simultaneously, P(a′1) represents the probability of a′1 occurring, P(a4|a′1) represents the conditional probability of a4 occurring under the premise of a′1 occurring, b5 represents the event of soft decoration maintenance, and p((a′1,a3)∩b5) represents the probability of a′1, a4 and b5 occurring simultaneously;

[0136] The tenant maintenance probability of the tenant is calculated based on the hard decoration maintenance probability corresponding to the tenant housing information and the soft decoration maintenance probability corresponding to the tenant housing information.

[0137] Optionally, the principle of calculating the tenant maintenance probability of the tenant based on the hard decoration maintenance probability corresponding to the tenant housing information and the soft decoration maintenance probability corresponding to the tenant housing information is similar to the principle of calculating the owner maintenance probability of the owner based on the hard decoration maintenance probability and the soft decoration maintenance probability, and will not be further elaborated here.

[0138] S3. Collect public equipment videos of users in the cell, identify device usage information of the public equipment in the public equipment videos, calculate a device update probability of the public equipment using the device usage information, and schedule the device resources using the device update probability to obtain a device scheduling result for the device resources.

[0139] The embodiment of the present invention identifies the device usage information of the public device in the public device video to use deep learning to detect whether the device is damaged.

[0140] In one embodiment of the present invention, the identifying of device usage information of the public device in the public device video includes: identifying, from the public device video, the device usage duration of the public device when the device has not been updated; calculating the device damage probability of the public device using a preset YOLOv5s network model; identifying the degree of device damage of the public device using the device damage probability; and using the device usage duration and the device damage degree as the device usage information.

[0141] It should be noted that the process of identifying the device usage duration when the public device is not updated from the public device video is distinguished by determining whether the device has been damaged artificially, and the principle of determining whether the device has been damaged artificially is similar to the principle of identifying abnormal human body posture features from the public device video described below.

[0142] Furthermore, in an embodiment of the present invention, the device update probability of the public device is calculated by using the device usage information, so as to identify abnormal human body posture features from the public device video.

[0143] Wherein, the device update probability refers to the probability of whether the device needs to be updated. The greater the degree of abnormal damage to the device, the greater the device update probability.

[0144] In an embodiment of the present invention, calculating the device update probability of the public device by using the device usage information includes: obtaining the public device video corresponding to the device usage information; extracting the user-device interaction feature between the community user and the public device from the public device video by using the following formula:

[0145]

[0146] Where B represents the user-device interaction feature, I i (x, y) represents the i-th video frame in the public device video, i~n represents the serial numbers of the first n video frames among the first n + 1 video frames, α represents the binarization threshold for separating the foreground area and the background area in the (n + 1)-th video frame, I n+1 (x, y) represents the (n + 1)-th video frame in the public device video, A n+1 (x, y) represents the foreground area in the (n + 1)-th video frame, (A n+1 (x, y)) x represents the gradient of the foreground area in the (n + 1)-th video frame in the horizontal direction, (A n+1 (x, y)) y represents the gradient of the foreground area in the (n + 1)-th video frame in the vertical direction, G(x, y) represents the gradient amplitude of the foreground area in the (n + 1)-th video frame, represents the ordinate value of the center point of the foreground area in the (n + 1)-th video frame, represents the minimum value of the ordinate values of the center points of the foreground areas in the i~n + 1 video frames, represents the ordinate value of the center point of the foreground area in the i-th video frame;

[0147] Obtain the device usage duration and the device damage degree in the device usage information; based on the device usage duration, the device damage degree, and the user-device interaction characteristics, use the following formula to calculate the device update probability of the public device:

[0148] f = sign(w T φ + b)

[0149] where f represents the device update probability, w represents the weight, b represents the bias, and φ represents a linear function composed of the device usage duration, the device damage degree, and the user-device interaction characteristics.

[0150] Among them, the user-device interaction characteristics refer to the human body posture characteristics when the user approaches the public device.

[0151] Furthermore, the embodiment of the present invention performs resource scheduling on the device resources by using the device update probability, so as to replace the device in time when device anomalies are detected through the device update probability.

[0152] In an embodiment of the present invention, the resource scheduling of the device resources by using the device update probability to obtain the device scheduling result of the device resources includes: using the device update probability to determine whether a device anomaly occurs in the public device; when a device anomaly occurs in the public device, use the following formula to calculate the device reliability of the public device:

[0153]

[0154] where Reliability represents the device reliability, β and γ represent constant parameters calculated based on the conditional probability when a device anomaly occurs in the public device, and t represents a time variable;

[0155] Based on the device reliability, use the following formula to calculate the device update time of the device resources:

[0156] Reliability ≥ δ

[0157]

[0158] where t max represents the device update time, δ represents the threshold of the device reliability, Reliability represents the device reliability, β and γ represent constant parameters calculated based on the conditional probability when a device anomaly occurs in the public device, and t represents a time variable;

[0159] Perform resource scheduling on the device resources within the device update time to obtain the device scheduling result of the device resources.

[0160] In another embodiment of the present invention, the resource scheduling of the device resources within the device update time to obtain the device scheduling result of the device resources includes: querying whether the device resources are in an idle state within the device update time; when the device resources are in an idle state within the device update time, identifying the current position of the device resources; determining whether the device resources are contended by other preset common devices; when the device resources are contended by other preset common devices, calculating the property-target distance and the property-other distance between the property area under the intelligent property management and the target common device and the other common device respectively; when the sum of the property-target distance and the property-other distance is the minimum value, storing the device resources at the property area to perform resource scheduling on the device resources and obtain the device scheduling result of the device resources; when the device resources are not contended by other preset common devices, calculating the property-current distance and the property-common distance between the property area under the intelligent property management and the current position and the target common device respectively; when the sum of the property-current distance and the property-common distance is the minimum value, storing the device resources at the property area to perform resource scheduling on the device resources and obtain the device scheduling result of the device resources; when the device resources are not in an idle state within the device update time, after delaying the device update time, returning to the step of querying whether the device resources are in an idle state within the device update time.

[0161] Optionally, the process of determining whether the device resources are contended by other preset common devices refers to the process of determining whether the device resources are contended by two target common devices at the same time. If device anomalies occur simultaneously in two target common devices, the phenomenon that the device resources are contended by two target common devices at the same time will occur.

[0162] It should be noted that the reason for storing the device resources at the property area to perform resource scheduling on the device resources is that the process of calculating the device update probability of the common device using the device usage information is a predictive process, that is, predicting whether the device will have an anomaly, rather than actually detecting a device anomaly. Therefore, the device that may have an anomaly cannot be immediately replaced with the device resources, but the location where the device resources are stored is placed near the device where the anomaly may occur, so that when the device has an anomaly subsequently, the abnormal device can be replaced with the device resources of the same type in a timely manner.

[0163] S4. Determine the data scheduling actions of the users in the cell for the data resources, where the data scheduling actions include modification, addition, deletion, and query. Collect the action influencing factors of the data scheduling actions, calculate the data scheduling probability of the data resources using the action influencing factors, and schedule the data resources using the data scheduling probability to obtain the data scheduling result of the data resources.

[0164] In the embodiments of the present invention, the action influencing factors include cell activities, holidays, age, housing information, etc.

[0165] Optionally, the principle of calculating the data scheduling probability of the data resources using the action influencing factors is similar to the principle of calculating the conditional probability in calculating the owner maintenance probability of the owner side using the housing information of the owner, and will not be elaborated further here.

[0166] S5. Use the human resource scheduling result, the equipment scheduling result, and the data scheduling result as the resource scheduling result under the intelligent property management.

[0167] It can be seen that in the embodiments of the present invention, human resources, equipment resources, and data resources under the intelligent property management are collected to summarize the schedulable resources under the intelligent property management. Further, in the embodiments of the present invention, the owner maintenance probability of the owner side is calculated using the housing information of the owner to predict the probability that the residence needs to be repaired by human resources based on the existing housing information of the owner through conditional probability. In the embodiments of the present invention, the equipment usage information of the public equipment in the public equipment video is identified to detect whether the equipment is damaged using deep learning. Further, in the embodiments of the present invention, the equipment update probability of the public equipment is calculated using the equipment usage information to identify abnormal human posture features from the public equipment video. Further, in the embodiments of the present invention, the equipment resources are scheduled using the equipment update probability to replace the equipment in a timely manner when equipment anomalies are detected through the equipment update probability. Therefore, a resource scheduling method and system under the intelligent property management based on deep learning proposed in the embodiments of the present invention can consider complex cell data for resource scheduling.

[0168] As Figure 2 shown, it is the functional module diagram of the resource scheduling system under the intelligent property management based on deep learning of the present invention.

[0169] The resource scheduling system 200 for implementing smart property management based on deep learning described in the present invention can be installed in an electronic device. Depending on the functionality implemented, the resource scheduling system for implementing smart property management based on deep learning can include a user segmentation module 201, a labor scheduling module 202, a device scheduling module 203, a data scheduling module 204, and a result determination module 205. A module, also referred to as a unit, is a series of computer program segments that can be executed by an electronic device processor and perform a fixed function. These modules are stored in the memory of the electronic device.

[0170] In the embodiment of the present invention, the functions of each module / unit are as follows:

[0171] The user segmentation module 201 is used to collect human resources, equipment resources, and data resources of the smart property and divide the users of the community under the smart property into owners and tenants;

[0172] The human resource scheduling module 202 is configured to collect the owner's housing information and the tenant's housing information, calculate the owner's maintenance probability of the owner using the owner's housing information, calculate the tenant's maintenance probability of the tenant using the tenant's housing information, and schedule the human resources using the owner's maintenance probability and the tenant's maintenance probability to obtain a human resource scheduling result.

[0173] The device scheduling module 203 is configured to collect public device videos of users in the cell, identify device usage information of the public devices in the public device videos, calculate device update probabilities of the public devices using the device usage information, and schedule the device resources using the device update probabilities to obtain device scheduling results for the device resources;

[0174] The data scheduling module 204 is configured to determine a data scheduling action of the cell user on the data resource, wherein the data scheduling action includes modification, addition, deletion, and query, collect action influencing factors of the data scheduling action, calculate a data scheduling probability of the data resource using the action influencing factors, schedule the data resource using the data scheduling probability, and obtain a data scheduling result for the data resource;

[0175] The result determination module 205 is used to use the manpower scheduling result, the equipment scheduling result, and the data scheduling result as the resource scheduling result of the smart property.

[0176] In detail, the modules in the resource scheduling system 200 for realizing smart property based on deep learning in the embodiment of the present invention are used in the same manner as above. Figure 1The same technical means as the resource scheduling method under intelligent property management implemented based on deep learning described in [reference] can be used, and the same technical effects can be achieved, which will not be elaborated here.

[0177] In addition, in each embodiment of the present invention, each functional module can be integrated into a processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware, or in the form of a combination of hardware and software functional modules.

[0178] For those skilled in the art, it is obvious that the present invention is not limited to the details of the above-mentioned exemplary embodiments, and without departing from the spirit or basic characteristics of the present invention, the present invention can be implemented in other specific forms.

[0179] Therefore, from any point of view, the embodiments should be regarded as exemplary and non-limiting. The scope of the present invention is defined by the appended claims rather than the above description. Therefore, all changes falling within the meaning and scope of the equivalent elements of the claims are intended to be included in the present invention. Any associated drawing marks in the claims should not be regarded as limiting the claims involved.

[0180] It should be noted that in this text, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements not only includes those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "comprising one..." does not exclude the existence of additional identical elements in the process, method, article or device comprising the said element.

[0181] The above are only specific embodiments of the present invention, enabling those skilled in the art to understand or implement the present invention. Various modifications to these embodiments will be obvious to those skilled in the art. The general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the broadest scope consistent with the principles and novel features disclosed herein.

Claims

1. A resource scheduling method based on deep learning for intelligent property management, characterized in that, The method includes: Collecting human resources, equipment resources, and data resources under intelligent property management, and dividing the community users under the intelligent property management into the owner side and the tenant side; Collecting the owner housing information of the owner side and the tenant housing information of the tenant side, calculating the owner maintenance probability of the owner side using the owner housing information, calculating the tenant maintenance probability of the tenant side using the tenant housing information, and scheduling the human resources using the owner maintenance probability and the tenant maintenance probability to obtain the human scheduling result of the human resources; Collecting the public equipment videos of the community users, identifying the equipment usage information of the public equipment in the public equipment videos, calculating the equipment update probability of the public equipment using the equipment usage information, and scheduling the equipment resources using the equipment update probability to obtain the equipment scheduling result of the equipment resources. Among them, calculating the equipment update probability of the public equipment using the equipment usage information includes: Obtaining the public equipment video corresponding to the equipment usage information; Using the following formula to extract the user-equipment interaction features between the community users and the public equipment from the public equipment video: Among them, B represents user-device interaction features, I i (x, y) represents the i-th video frame in the public device video, i~n represents the serial numbers of the first n video frames among the first n + 1 video frames, α represents the binarization threshold for separating the foreground area and the background area in the (n + 1)-th video frame, I n+1 (x, y) represents the (n + 1)-th video frame in the public device video, A n+1 (x, y) represents the foreground area in the (n + 1)-th video frame, (A n+1 (x, y)) x represents the gradient of the foreground area in the (n + 1)-th video frame in the horizontal direction, (A n+1 (x, y)) y represents the gradient of the foreground area in the (n + 1)-th video frame in the vertical direction, G(x, y) represents the gradient magnitude of the foreground area in the (n + 1)-th video frame, represents the ordinate value of the center point of the foreground area in the (n + 1)-th video frame, represents the minimum value of the ordinate values of the center points of the foreground areas in the i~n + 1 video frames, represents the ordinate value of the center point of the foreground area in the i-th video frame; Obtaining the equipment usage duration and equipment damage degree in the equipment usage information; Based on the equipment usage duration, the equipment damage degree, and the user-equipment interaction features, using the following formula to calculate the equipment update probability of the public equipment: f = sign(w T φ + b) Where f represents the equipment update probability, w represents the weight, b represents the bias, and φ represents a linear function composed of the equipment usage duration, the equipment damage degree, and the user-equipment interaction features; Determining the data scheduling actions of the community users for the data resources, where the data scheduling actions include modification, addition, deletion, and query, collecting the action influencing factors of the data scheduling actions, calculating the data scheduling probability of the data resources using the action influencing factors, and scheduling the data resources using the data scheduling probability to obtain the data scheduling result of the data resources; Taking the human scheduling result, the equipment scheduling result, and the data scheduling result as the resource scheduling result under the intelligent property management.

2. The method according to claim 1, characterized in that, The collecting of the human resources, equipment resources, and data resources under the intelligent property management includes: Collecting the human data, equipment data, and data data under the intelligent property management; Based on the human data, calculating the human resources using the following formula: Among them, represents the human resources belonging to the i-th property area, and m represents the total number of property human resources in the human data that belong to the i-th property area and are in the idle state rest. represents the j-th property human resource in the human data that belongs to the i-th property area and is in the idle state rest; Based on the equipment data, calculating the equipment resources using the following formula: Among them, represents the device resources of the k-th type, and S represents the total number of devices of the k-th type in the device data. represents the s-th device of the k-th type in the device data; Based on the data data, calculating the data resources using the following formula: Among them, represents the data resources belonging to the i-th property area, and L represents the total number of community users in the data belonging to the i-th property area. represents the amount of data of the l-th community user in the data belonging to the i-th property area.

3. The method according to claim 1, wherein The calculating of the owner maintenance probability of the owner side using the owner housing information includes: Distinguishing the hard decoration housing information and the soft decoration housing information in the owner housing information; Calculating the hard decoration maintenance probability of the hard decoration housing information using the following formula: p(a1,a2,a3)=P(a1)P(a2∣a1)P(a3∣a1,a2) Among them, p(b1|a1,a2,a3) represents the hard decoration maintenance probability, a1 represents the housing construction duration in the hard decoration housing information, a2 represents the housing damage degree in the hard decoration housing information, a3 represents the owner category in the hard decoration housing information, p(a1,a2,a3) represents the probability that a1, a2, and a3 occur simultaneously, P(a1) represents the probability that a1 occurs, P(a2∣a1) represents the conditional probability that a2 occurs on the premise that a1 occurs, P(a3∣a1,a2) represents the conditional probability that a3 occurs on the premise that a1 and a2 occur, b1 represents the event of hard decoration maintenance, and p((a1,a2,a3)∩b1) represents the probability that a1, a2, a3 and b1 occur simultaneously; Calculate the soft decoration maintenance probability of the soft decoration housing information by using the following formula: p(a′1,a3) = P(a′1)P(a3∣a′1) Among them, p(b2|a′1,a3) represents the soft decoration maintenance probability, a′1 represents the housing decoration duration in the soft decoration housing information, a3 represents the owner category in the soft decoration housing information, p(a′1,a3) represents the probability that a′1 and a3 occur simultaneously, P(a′1) represents the probability that a′1 occurs, P(a3∣a′1) represents the conditional probability that a3 occurs on the premise that a′1 occurs, b2 represents the event of soft decoration maintenance, and p((a′1,a3)∩b2) represents the probability that a′1, a3 and b2 occur simultaneously; Calculate the owner maintenance probability of the owner side based on the hard decoration maintenance probability and the soft decoration maintenance probability.

4. The method according to claim 3, wherein The calculating the owner maintenance probability of the owner side based on the hard decoration maintenance probability and the soft decoration maintenance probability includes: Calculate the owner maintenance probability of the owner side by using the following formula based on the hard decoration maintenance probability and the soft decoration maintenance probability: P(b1) = p(b1|a1,a2,a3) P(b2) = p(b2|a′1,a3) Among them, p(b3|b1,b2) represents the owner maintenance probability, p(b1|a1,a2,a3) represents the hard decoration maintenance probability, p(b2|a′1,a3) represents the soft decoration maintenance probability, p(b1,b2,b3) represents the probability that b1, b2, and b3 occur simultaneously, p(b1,b2) represents the probability that b1 and b2 occur simultaneously, and P(b2∣b1) represents the conditional probability that b2 occurs on the premise that b1 occurs.

5. The method according to claim 1, wherein The calculating the tenant maintenance probability of the tenant side by using the tenant housing information includes: Calculate the hard decoration maintenance probability corresponding to the tenant housing information by using the following formula: p(a1,a2,a4) = P(a1)P(a2∣a1)P(a4∣a1,a2) Among them, p(b4|a1,a2,a4) represents the probability of hard decoration maintenance corresponding to the tenant housing information, a1 represents the construction duration of the residence in the hard decoration housing information, a2 represents the degree of damage of the residence in the hard decoration housing information, a4 represents the tenant category in the tenant housing information, p(a1,a2,a4) represents the probability that a1, a2, and a4 occur simultaneously, P(a1) represents the probability that a1 occurs, P(a2∣a1) represents the conditional probability that a2 occurs on the premise that a1 occurs, P(a4∣a1,a2) represents the conditional probability that a4 occurs on the premise that a1 and a2 occur, b4 represents the event of hard decoration maintenance corresponding to the tenant housing information, and p((a1,a2,a4)∩b4) represents the probability that a1, a2, a4 and b4 occur simultaneously; Calculate the probability of soft decoration maintenance corresponding to the tenant housing information by using the following formula: p(a′1,a4)=P(a′1)P(a4∣a′1) Among them, p(b5|a′1,a3) represents the probability of soft decoration maintenance corresponding to the tenant housing information, a′1 represents the decoration duration of the residence in the soft decoration housing information, a4 represents the tenant category in the tenant housing information, p(a′1,a4) represents the probability that a′1 and a4 occur simultaneously, P(a′1) represents the probability that a′1 occurs, P(a4∣a′1) represents the conditional probability that a4 occurs on the premise that a′1 occurs, b5 represents the event of soft decoration maintenance, and p((a′1,a3)∩b5) represents the probability that a′1, a4 and b5 occur simultaneously; Calculate the tenant maintenance probability of the tenant party based on the probability of hard decoration maintenance corresponding to the tenant housing information and the probability of soft decoration maintenance corresponding to the tenant housing information.

6. The method according to claim 1, wherein The identification of the equipment usage information of the public equipment in the public equipment video includes: Identify the equipment usage duration when the public equipment has not been updated from the public equipment video; Calculate the equipment damage probability of the public equipment by using a preset YOLOv5s network model; Identify the degree of equipment damage of the public equipment by using the equipment damage probability; Use the equipment usage duration and the degree of equipment damage as the equipment usage information.

7. The method according to claim 1, characterized in that, The use of the equipment update probability to perform resource scheduling on the equipment resources to obtain the equipment scheduling result of the equipment resources includes: Use the equipment update probability to determine whether the public equipment has equipment anomalies; When the public equipment has equipment anomalies, calculate the equipment reliability of the public equipment by using the following formula: Among them, Reliability represents equipment reliability, β, γ represent constant parameters calculated based on the conditional probability when the public equipment has equipment anomalies, and t represents a time variable; Based on the equipment reliability, calculate the equipment update time of the equipment resources by using the following formula: Reliability≥δ where t max represents the device update time, δ represents the threshold of device reliability, Reliability represents the device reliability, β and γ represent constant parameters calculated based on the conditional probability when a device anomaly occurs in the public device, and t represents the time variable; Perform resource scheduling on the equipment resources within the equipment update time to obtain the equipment scheduling result of the equipment resources.

8. The method according to claim 1, wherein Performing resource scheduling on the device resources within the device update time to obtain a device scheduling result of the device resources, including: Querying whether the device resources are in an idle state within the device update time; When the device resources are in an idle state within the device update time, identifying the current location of the device resources; Judging whether the device resources are contended by other preset public devices; When the device resources are contended by other preset public devices, calculating the property-target distance and the property-other distance between the property area under the intelligent property and the target public device and the other public device respectively; When the sum of the property-target distance and the property-other distance is the minimum value, storing the device resources at the property area to perform resource scheduling on the device resources and obtain a device scheduling result of the device resources; When the device resources are not contended by other preset public devices, calculating the property-current distance and the property-public distance between the property area under the intelligent property and the current location and the target public device respectively; When the sum of the property-current distance and the property-public distance is the minimum value, storing the device resources at the property area to perform resource scheduling on the device resources and obtain a device scheduling result of the device resources; When the device resources are not in an idle state within the device update time, after delaying the device update time, returning to the step of querying whether the device resources are in an idle state within the device update time above.

9. A resource scheduling system based on deep learning for intelligent property management, characterized in that, The system includes: A user division module, configured to collect human resources, device resources, and material resources under the intelligent property, and divide the community users under the intelligent property into an owner party and a tenant party; A human resource scheduling module, configured to collect the owner housing information of the owner party and the tenant housing information of the tenant party, calculate the owner maintenance probability of the owner party using the owner housing information, calculate the tenant maintenance probability of the tenant party using the tenant housing information, and schedule the human resources using the owner maintenance probability and the tenant maintenance probability to obtain a human resource scheduling result of the human resources; A device scheduling module, configured to collect public device videos of the community users, identify device usage information of the public devices in the public device videos, calculate the device update probability of the public devices using the device usage information, and perform resource scheduling on the device resources using the device update probability to obtain a device scheduling result of the device resources. Among them, calculating the device update probability of the public devices using the device usage information includes: Obtaining the public device video corresponding to the device usage information; Extracting the user-device interaction feature between the community users and the public devices from the public device video using the following formula: Among them, B represents user-device interaction features, I i (x, y) represents the i-th video frame in the public device video, i~n represents the serial numbers of the first n video frames among the first n + 1 video frames, α represents the binary threshold for separating the foreground region and the background region in the (n + 1)-th video frame, I n+1 (x, y) represents the (n + 1)-th video frame in the public device video, A n+1 (x, y) represents the foreground region in the (n + 1)-th video frame, (A n+1 (x, y)) x represents the gradient of the foreground region in the (n + 1)-th video frame in the horizontal direction, (A n+1 (x, y)) y represents the gradient of the foreground region in the (n + 1)-th video frame in the vertical direction, G(x, y) represents the gradient amplitude of the foreground region in the (n + 1)-th video frame, represents the ordinate value of the center point of the foreground region in the (n + 1)-th video frame, represents the minimum value of the ordinate values of the center points of the foreground regions in the i~n + 1 video frames, represents the ordinate value of the center point of the foreground region in the i-th video frame; Obtaining the device usage duration and the device damage degree in the device usage information; Calculating the device update probability of the public device using the following formula based on the device usage duration, the device damage degree, and the user-device interaction feature: f = sign(w T φ + b) Among them, f represents the device update probability, w represents the weight, b represents the bias, and φ represents a linear function composed of the device usage duration, the device damage degree, and the user-device interaction characteristics; A data scheduling module, configured to determine the data scheduling actions of the cell users for the data resources, where the data scheduling actions include modification, addition, deletion, and query, collect the action influencing factors of the data scheduling actions, calculate the data scheduling probability of the data resources using the action influencing factors, and schedule the data resources using the data scheduling probability to obtain the data scheduling result of the data resources; A result determination module, configured to use the personnel scheduling result, the device scheduling result, and the data scheduling result as the resource scheduling result under the intelligent property management.

Images