Public facility management and control method and management and control system based on Internet of Things, and storage medium

Abstract

The invention discloses a public facility management and control method and system based on the Internet of Things and a storage medium, and the method comprises the steps: determining the type of a to-be-managed and controlled public facility according to a type identifier carried by a management and control instruction when the server of the Internet of Things detects the management and control instruction; if the shared bikes are the shared bikes, determining the number of the to-be-released bikes of the shared bikes in each preset release point of the preset area according to the circulation data of the shared bikes in the preset area in the preset time period; according to the circulation data and the number of the to-be-delivered bicycles, scheduling information of each preset delivery point is generated to schedule the shared bicycles in the preset area; and determining maintenance information of the sports facilities according to the facility images of the sports facilities in the preset area at the plurality of visual angles if the sports facilities belong to the sports facility type, and managing and controlling the sports facilities in the preset area. According to the invention, for various communal facilities, respective related information is obtained through the Internet of Things server for corresponding management and control maintenance, and accurate management and control of the communal facilities are realized.

Description

Technical Field

[0001] The present invention relates to the technical field of Internet of Things, and in particular to a public facility management and control method, a management and control system and a storage medium based on the Internet of Things. Background Art

[0002] Public facilities are various public and service facilities provided to the public, which may include public health facilities, public sports facilities, public cultural facilities, public transportation facilities, public greening facilities, etc. With the improvement of residents' economic level, there are more and more types of public facilities, and the areas covered are becoming more and more extensive. Although the increase in types and areas can bring more convenience to the public, it is accompanied by the management and control of numerous public facilities. For example, for shared bicycles as public transportation facilities, although the areas where they are deployed are increasing, there may be a situation where the number of bicycles deployed in one area is insufficient and the number of bicycles deployed in another area is excessive, which will affect the convenience of use for the public on the one hand, and may cause road traffic congestion due to accumulation on the other hand. Or for public sports facilities, damage may occur and fail to be discovered in time, which will lead to safety hazards. Therefore, how to accurately control the growing number of public facilities is a technical problem that needs to be solved urgently. Summary of the invention

[0003] The main purpose of the present invention is to provide a public facility management and control method, a management and control system and a storage medium based on the Internet of Things, aiming to solve the technical problem of how to perform refined management and control of public facilities.

[0004] To achieve the above object, the present invention provides a public facility management and control method based on the Internet of Things, which is applied to an Internet of Things server. The public facility management and control based on the Internet of Things includes:

[0005] When the IoT server detects a control instruction triggered based on a preset control cycle, it identifies the type identifier carried by the control instruction, and determines the type of the public facility to be controlled according to the type identifier, where different types of public facilities correspond to different preset control cycles;

[0006] If the type of the public facility to be controlled is a shared bicycle type, then the flow data of the shared bicycles in the preset area within the preset time period is obtained, and according to the flow data, the number of shared bicycles to be placed in each preset placement point in the preset area is determined;

[0007] Generate scheduling information for each preset delivery point according to the flow data and the number of bicycles to be delivered, and schedule the shared bicycles in the preset area according to the scheduling information to manage and control the shared bicycles in the preset area;

[0008] If the type of the public facility to be controlled is a sports facility, facility images of the sports facilities in a preset area from multiple perspectives are obtained, and maintenance information of the sports facilities is determined based on each of the facility images, and the sports facilities in the preset area are controlled based on the maintenance information.

[0009] Preferably, the step of generating the scheduling information of each preset delivery point according to the flow data and the number of bicycles to be delivered includes:

[0010] Determine the final destination of the shared bicycle within the preset time period according to the identification time of the location identification carried by the circulation data, and count the number of shared bicycles at each final destination to obtain the number of existing bicycles at each preset delivery point;

[0011] For each of the preset delivery points, determining the relationship between the number of existing bicycles and the number of bicycles to be delivered;

[0012] If the size relationship is that the number of existing bicycles is greater than the number of bicycles to be delivered, generating information to be called out for the preset delivery point, and generating the scheduling information according to the information to be called out and the adjacent delivery points corresponding to the preset delivery point;

[0013] If the size relationship is that the number of existing bicycles is less than the number of bicycles to be deployed, the pending transfer information is generated for the preset delivery point, and the scheduling information is generated based on the pending transfer information and the adjacent delivery points corresponding to the preset delivery point.

[0014] Preferably, the step of generating the scheduling information according to the information to be called out and the adjacent delivery points corresponding to the preset delivery points includes:

[0015] Find out the first-category adjacent delivery points that have not generated scheduling information from each of the adjacent delivery points, and find out the second-category adjacent delivery points that carry information to be transferred from the first-category adjacent delivery points;

[0016] Determine the position distances between the preset delivery point and each of the second-category adjacent delivery points, and determine a target adjacent delivery point from each of the second-category adjacent delivery points according to each of the position distances;

[0017] The information to be called out and the target adjacent delivery point are generated as the scheduling information.

[0018] Preferably, the step of generating the scheduling information of each preset delivery point according to the flow data and the number of bicycles to be delivered comprises:

[0019] For each of the preset delivery points, obtaining the riding change parameters and optimization evaluation coefficients of the preset delivery point in multiple historical preset time periods, as well as the weather parameters in the next preset time period;

[0020] Find the historical damage ratio of shared bicycles in the preset area, and update the quantity information in the scheduling information according to the historical damage ratio, riding change parameters, weather parameters and optimization evaluation coefficient to obtain the updated scheduling information of the preset delivery point, wherein the update calculation formula is:

[0021]

[0022] Among them, yi represents the quantity information of the i-th preset delivery point, represents the weather parameters of the i-th preset delivery point, θ mi represents the optimization evaluation coefficient of the ith preset delivery point in the mth historical preset period, ε represents the historical damage ratio of the ith preset delivery point, τ mi It represents the riding change parameter of the i-th preset delivery point in the m-th historical preset time period, and n represents the number of historical preset time periods.

[0023] Preferably, the step of determining the number of shared bicycles to be placed in each preset placement point in the preset area according to the flow data includes:

[0024] Identify the location identifiers carried by each item of the flow data, and count the number of identical location identifiers to obtain multiple statistical results;

[0025] According to the correspondence between the position identifier of each statistical result and each preset delivery point, the number of shared bicycles to be delivered at each preset delivery point is determined.

[0026] Preferably, the step of determining the maintenance information of the sports facilities according to each of the facility images comprises:

[0027] For each of the facility images, obtaining three primary color values ​​of pixels in the facility image, and converting the three primary color values ​​of each of the pixels into grayscale values ​​to obtain a grayscale image corresponding to the facility image;

[0028] Extracting a contour image corresponding to the sports facility in the grayscale image according to the grayscale value of each pixel in the grayscale image, and obtaining contour images of the sports facility at multiple viewing angles;

[0029] Transmitting each of the contour images to a preset recognition model, and processing a pixel matrix formed by the grayscale values ​​of each pixel in each of the contour images based on the first convolution kernel and the first pooling layer parameters in the first convolution layer of the preset recognition model to obtain a type matrix for determining the type of the item to which the sports facility belongs;

[0030] Based on the second convolution kernel and the second pooling layer parameters in the second convolution layer corresponding to the item type in the preset recognition model, the type matrix is ​​calculated to obtain the damage probability, and it is judged whether the current sports facility is damaged according to the damage probability, wherein the calculation formula is:

[0031]

[0032] Where p(o) represents the damage probability value, k2 represents the second convolution kernel parameter, k1 represents the first convolution kernel parameter, and W i (i 1 , j 1 ) represents the pixel matrix of the i-th contour image. 1 row, j 1 The value of the column, S1(i 1 ,j 1 ) represents the i-th convolution kernel in the first convolution kernel 1 row, j 1 The value of the column, α represents the step size corresponding to the first convolution kernel, m k1 represents the first pooling layer parameters, Indicates the i-th 2 row, j 2 The value of the column, S2(i 2 , j 2 ) represents the i-th convolution kernel in the second convolution kernel 2 row, j 2 The value of the column, β represents the step size corresponding to the second convolution kernel, m k2 Represents the parameters of the second pooling layer.

[0033] If there is damage, the repair information is generated according to the damage type corresponding to the damage probability.

[0034] Preferably, if damage exists, the step of generating the repair information according to the damage type corresponding to the damage probability comprises:

[0035] Identify the grayscale image of the damaged sports facilities as training data, and update and train the preset recognition model according to the training data to obtain a training result;

[0036] The loss function value of the preset recognition model is calculated according to the training result, and the calculation formula is:

[0037]

[0038] Where L represents the loss function, A represents the overlap area ratio between the first minimum bounding rectangle corresponding to the true outline of the sports facilities in the grayscale image and the second minimum bounding rectangle in the training result, and d z The center distance between the center point of the first minimum enclosing rectangle and the center point of the second minimum enclosing rectangle, d s represents the distance between the farthest edges of the first minimum circumscribed rectangle and the second minimum circumscribed rectangle, a represents a preset formula parameter, w1 and h1 represent the width and length of the first minimum circumscribed rectangle, and w2 and h2 represent the width and length of the second minimum circumscribed rectangle;

[0039] According to the magnitude relationship between the loss function value and the preset threshold, the grayscale image recognition weight matrix of the preset recognition model is updated.

[0040] Preferably, the step of updating the grayscale image recognition weight matrix of the preset recognition model according to the magnitude relationship between the loss function value and the preset threshold value comprises:

[0041] Determine whether the magnitude relationship is that the loss function value is greater than a preset threshold;

[0042] If the loss function value is greater than a preset threshold, the grayscale image recognition weight matrix of the preset recognition model is updated based on a preset update formula, so as to perform the step of updating and training the preset recognition model according to the training data based on the updated grayscale image recognition weight matrix to obtain a training result;

[0043] Among them, the preset update formula is:

[0044]

[0045] Among them, V(i 3 ,j 3 ) represents the grayscale image recognition weight matrix, x(i 3 ,j 3 ) represents the i-th grayscale image recognition weight matrix 3 row, j 3 The value of the column, k3 represents the number of rows and columns of the grayscale image recognition weight matrix, Z(i 3 , j 3 ) represents the preset bias matrix.

[0046] Furthermore, to achieve the above-mentioned purpose, the present invention also provides a public facility management and control system based on the Internet of Things, wherein the public facility management and control system based on the Internet of Things includes an Internet of Things server, wherein the Internet of Things server is provided with a storage device, a processor, a communication bus, and a control program stored on the storage device:

[0047] The communication bus is used to realize the connection and communication between the processor and the storage;

[0048] The processor is used to execute the control program to implement the steps of the public facility management and control method based on the Internet of Things as described above.

[0049] Furthermore, to achieve the above-mentioned purpose, the present invention also provides a storage medium, on which a control program is stored, and when the control program is executed by a processor, the steps of the public facility management and control method based on the Internet of Things as described above are implemented.

[0050] The present invention is based on a public facility control method, a control system and a storage medium of the Internet of Things. The method is applied to an Internet of Things server. Different preset control cycles are pre-set for different types of public facilities. When the Internet of Things server detects a control instruction triggered based on the preset control cycle, it identifies the type identifier carried by the control instruction, and determines the type of the public facility to be controlled according to the type identifier; if it is determined through identification that the type of the public facility to be controlled is a shared bicycle type, then the flow data of all shared bicycles in a preset area within a preset time period is obtained, and based on the obtained flow data, the number of bicycles to be released at each preset delivery point in the preset area is determined; then, based on the flow data and the number of bicycles to be released, the scheduling information of each preset delivery point is generated to schedule the shared bicycles in the preset area; if it is determined through identification that the type of the public facility to be controlled is a sports facility type, then the facility images of the sports facilities in the preset area at multiple perspectives are obtained, and the maintenance information of the sports facilities is determined based on each facility image to control and maintain the sports facilities. Among them, the number of bicycles to be put into each preset delivery point generated based on the circulation data reflects the actual demand for shared bicycles at each delivery point. Dispatching shared bicycles based on the generated dispatch information can make shared bicycles provide convenience to the public to the greatest extent while avoiding traffic congestion; for sports facilities, by collecting facility images from multiple perspectives for identification, forming maintenance information for management and maintenance, safety hazards caused by damage to sports facilities can be discovered in time. In this way, by setting different preset control cycles for various public facilities in the preset area, the relevant information is obtained through the Internet of Things server every time the respective control cycle is reached to carry out corresponding control and maintenance of the public facilities, thus realizing accurate control of public facilities. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1This is a flow chart of a first embodiment of a method for public facility management and control based on the Internet of Things of the present invention;

[0052] Figure 2 This is a flow chart of a second embodiment of a method for public facility management and control based on the Internet of Things of the present invention;

[0053] Figure 3 This is a flow chart of a third embodiment of a method for controlling public facilities based on the Internet of Things according to the present invention;

[0054] Figure 4 It is a structural diagram of the hardware operating environment involved in an embodiment of a public facility management and control system based on the Internet of Things of the present invention.

[0055] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0056] 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.

[0057] The present invention provides a public facility management and control method based on the Internet of Things, which is applied to the Internet of Things server. Figure 1 , Figure 1 It is a flow chart of the first embodiment of the public facility management and control method based on the Internet of Things of the present invention.

[0058] The embodiment of the present invention provides an embodiment of a method for controlling public facilities based on the Internet of Things. It should be noted that although a logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than that here. Specifically, the method for controlling public facilities based on the Internet of Things in this embodiment includes:

[0059] Step S10, when the IoT server detects a control instruction triggered based on a preset control cycle, it identifies the type identifier carried by the control instruction, and determines the type of public facilities to be controlled based on the type identifier, where different types of public facilities correspond to different preset control cycles.

[0060] The public facility control method based on the Internet of Things in this embodiment is applied to the Internet of Things server, and different preset control cycles are set in advance for different types of public facilities. Every time a preset control cycle of a certain type of public facility is detected, the relevant information of the type of public facility is obtained through the Internet of Things server to carry out refined control. Specifically, it can include accurate scheduling of shared bicycles and damage identification of sports facilities.

[0061] Among them, the preset control cycle of shared bicycles and sports facilities can be set according to their respective characteristics. For example, for shared bicycles, it can be set according to their dependence on changes in population mobility, and for sports facilities, it can be set according to their damage cycle affected by climate factors and frequency of use. When the detection module in the Internet of Things server detects that the preset control cycle of a public facility has been reached, it triggers a control instruction indicating that the public facility needs to be controlled and maintained according to the preset control cycle. For example, if the preset control cycle of a shared bicycle is set to A, then a control instruction for controlling the contributed bicycle is triggered once every preset control cycle A.

[0062] Furthermore, after detecting the control instruction triggered based on the preset control cycle, the IoT server reads the type identifier carried in the control instruction for identification. The type identifier comes from the public facility and indicates which public facility needs to be controlled when it reaches the preset control cycle. The IoT server can determine the type of public facility to be controlled by identifying the type identifier.

[0063] Step S20: If the type of the public facility to be controlled is a shared bicycle type, the flow data of the shared bicycles in the preset area within the preset time period is obtained, and based on the flow data, the number of shared bicycles to be placed in each preset placement point in the preset area is determined.

[0064] Furthermore, if it is determined through identification that the type of public facilities to be controlled is a shared bicycle type, the circulation data of all shared bicycles in the preset area within the preset time period is obtained. The preset area is a pre-set area indicating that shared bicycles need to be controlled at the same time, for example, a district under a certain city. The preset time period is a time period with a high turnover rate of shared bicycles on each date within the preset control cycle, such as the morning rush hour and evening rush hour on weekdays. The circulation data represents the number of circulations and circulation positions of each shared bicycle within the preset time period. For example, from position A to position B and then to position C, the circulation data formed by the shared bicycle is 2 circulations and 3 positions.

[0065] Furthermore, the circulation data indicates the demand for shared bicycles in various places. For example, the circulation data of the above two circulations and three positions indicates that at least one shared bicycle is needed at both position A and position B. In addition, considering that shared bicycles cannot be parked at will, the road sections used for parking shared bicycles in road traffic are used as preset delivery points. The number of circulations and circulation positions of each shared bicycle contained in the circulation data are used to determine the number of shared bicycles required in each preset delivery point in the preset area, that is, the number of bicycles to be delivered. Specifically, the step of determining the number of shared bicycles to be delivered in each preset delivery point in the preset area based on the circulation data includes:

[0066] Step S21, identifying the location identifiers carried by each item of the circulating data, and counting the number of identical location identifiers to obtain multiple statistical results;

[0067] Step S22, determining the number of shared bicycles to be placed at each preset delivery point according to the correspondence between the position identifier of each statistical result and each preset delivery point.

[0068] Furthermore, a piece of circulation data is formed by the circulation of a shared bicycle within a preset time period, and carries the location identifiers of all the locations where the shared bicycle is parked within the preset time period. The Internet of Things server identifies this type of location identifier, and accumulates and counts the number of location identifiers identified as the same, to obtain multiple statistical results. It should be noted that for a piece of circulation data, the parking location within the preset time period may be repeated, so that the circulation data carries repeated location identifiers. For example, the circulation route of a shared bicycle is from position A to position B to position C and then to position A. At this time, the repeated location identifier needs to be counted as the same location identifier, and the number of location identifiers corresponding to position A is counted as 2, which is used to represent the demand for shared bicycles at the parking location corresponding to the location identifier.

[0069] Furthermore, after counting the location identifiers carried in each circulation data, multiple statistical results are obtained. All the location identifiers in a single statistical result are the same, indicating the same location where the shared bicycles are parked. The number of location identifiers in the statistical result indicates the demand for shared bicycles at that location. In addition, each preset delivery point in the preset area also carries a delivery point identifier indicating its respective location, so the location identifier of the statistical result can be compared with each delivery point identifier to find a delivery point identifier that matches the location identifier. The preset delivery point with the matching delivery point identifier is located at the location indicated by the location identifier, so the number of location identifiers in the statistical result can be determined as the number of shared bicycles to be delivered for the preset delivery point.

[0070] Step S30, generating scheduling information for each preset delivery point according to the flow data and the number of bicycles to be delivered, and scheduling the shared bicycles in the preset area according to the scheduling information to manage and control the shared bicycles in the preset area.

[0071] Understandably, a shared bicycle will eventually stay at a certain location within a preset time period. This location can be reflected by the last location identifier carried by the circulation data, corresponding to a preset delivery point. Therefore, for the number of bicycles to be delivered required by each preset delivery point, the number of shared bicycles actually required to be delivered at the preset delivery point will be reduced due to the final stay of the shared bicycles in the preset time period. For example, it is determined by statistics that the number of bicycles to be delivered at a preset delivery point is 100, but 50 shared bicycles finally stay at the preset delivery point within the preset time period, then the actual number of shared bicycles to be delivered at the preset delivery point is 50. Therefore, it is necessary to combine the final parking location of the shared bicycle reflected in the circulation data with the number of bicycles to be delivered at each preset delivery point to generate the scheduling information of each preset delivery point. Then, the shared bicycles in the preset area are scheduled according to the scheduling information to achieve the management and control of shared bicycles.

[0072] It should be noted that the dispatch information of each preset delivery point includes whether the preset delivery point needs to transfer in or out shared bicycles, and the number of transferred in or out. When scheduling based on this type of dispatch information, the preset delivery point that needs to be transferred out can be preferentially selected as the starting point, and the transfer in and out status of other preset delivery points adjacent to it can be determined. The transfer in and out routes are set according to the transfer in and out status, so that the maintenance personnel of the shared bicycles can dispatch the shared bicycles in the preset area according to the transfer in and out routes, avoiding detours and saving dispatch time.

[0073] Step S40: If the type of the public facility to be controlled is a sports facility, obtain facility images of the sports facilities in a preset area from multiple perspectives, determine maintenance information of the sports facilities based on each of the facility images, and control the sports facilities in the preset area based on the maintenance information.

[0074] Furthermore, if the physical network server determines that the type of the public facility to be controlled is a sports facility through type identification, the facility images of all sports facilities in the preset area at various viewing angles are obtained. The sports facilities are preferably fitness equipment in the park, such as a waist twister, a traction device, parallel bars, etc. In addition, the multi-view facility images can be taken by a camera in the park that is connected to the IoT server, or a cruise robot or drone can be set up to take pictures.

[0075] Furthermore, after obtaining facility images of each sports facility from multiple perspectives, the facility images from each perspective can be identified, and the facility images from each perspective can be used to determine whether the sports facility is damaged, generate corresponding maintenance information, and manage the sports facility based on the maintenance information. If it is identified that the sports facility is not damaged in the facility images from each perspective, maintenance information that does not require maintenance is generated. On the contrary, if there is damage in the facility image from any perspective, the damage location and damage type are jointly generated as maintenance information that requires maintenance, and prompt information is output for the maintenance information to prompt the sports facility to be repaired, thereby achieving management and control of the sports facility.

[0076] The public facility control method based on the Internet of Things implemented in this embodiment pre-sets different preset control cycles for different types of public facilities. When the Internet of Things server detects a control instruction triggered based on the preset control cycle, it identifies the type identifier carried by the control instruction and determines the type of the public facility to be controlled based on the type identifier; if it is determined through identification that the type of the public facility to be controlled is a shared bicycle, the flow data of all shared bicycles in a preset area within a preset time period is obtained, and based on the obtained flow data, the number of bicycles to be released at each preset delivery point in the preset area is determined; then, based on the flow data and the number of bicycles to be released, the scheduling information of each preset delivery point is generated to schedule the shared bicycles in the preset area; if it is determined through identification that the type of the public facility to be controlled is a sports facility, the facility images of the sports facilities in the preset area from multiple perspectives are obtained, and the maintenance information of the sports facilities is determined based on each facility image to control and maintain the sports facilities. Among them, the number of bicycles to be put into each preset delivery point generated based on the circulation data reflects the actual demand for shared bicycles at each delivery point. Dispatching shared bicycles based on the generated dispatch information can make shared bicycles provide convenience to the public to the greatest extent while avoiding traffic congestion; for sports facilities, by collecting facility images from multiple perspectives for identification, forming maintenance information for management and maintenance, safety hazards caused by damage to sports facilities can be discovered in time. In this way, by setting different preset control cycles for various public facilities in the preset area, the relevant information is obtained through the Internet of Things server every time the respective control cycle is reached to carry out corresponding control and maintenance of the public facilities, thus realizing accurate control of public facilities.

[0077] For further information, please refer to Figure 2 Based on the first embodiment of the public facility management and control method based on the Internet of Things of the present invention, a second embodiment of the public facility management and control method based on the Internet of Things of the present invention is proposed.

[0078] The difference between the second embodiment of the public facility control method based on the Internet of Things and the first embodiment of the public facility control method based on the Internet of Things is that the step of generating the scheduling information of each preset delivery point according to the flow data and the number of bicycles to be delivered includes:

[0079] Step S31, determining the final destination of the shared bicycle within the preset time period according to the identification time of the location identification carried by the circulation data, and counting the number of shared bicycles at each of the final destinations to obtain the number of existing bicycles at each of the preset delivery points;

[0080] Step S32, for each of the preset delivery points, determining the relationship between the number of existing bicycles and the number of bicycles to be delivered;

[0081] Step S33, if the size relationship is that the number of existing bicycles is greater than the number of bicycles to be delivered, generating information to be called out for the preset delivery point, and generating the scheduling information according to the information to be called out and the adjacent delivery points corresponding to the preset delivery point;

[0082] Step S34: If the size relationship is that the number of existing bicycles is less than the number of bicycles to be delivered, then information to be transferred in is generated for the preset delivery point, and the scheduling information is generated based on the information to be transferred in and the adjacent delivery points corresponding to the preset delivery point.

[0083] Furthermore, the circulation data carries multiple location identifiers, and the identification time corresponding to each location identifier is different, indicating that the shared bicycle is parked at different locations at different times within the preset time period. In order to determine the final parking location of the shared bicycle in the preset time period, the identification time of the location identifier carried by each piece of circulation data is obtained, the latest time is determined, and the location identifier with this time is searched. The location represented by the location identifier is the location where the shared bicycle is finally parked in the preset time period, and it is determined as the final destination of the shared bicycle in the preset time period. Then, the number of shared bicycles in each final destination is counted. This statistics can be achieved by finding the target circulation data of the final destination whose location identifier at the latest time in each circulation data is found, and the number of target circulation data is counted. After the search and statistics of each circulation data are completed, the number of existing bicycles at each preset delivery point can be obtained. Among them, in the case where the preset delivery point is not the final destination, the number of existing bicycles at the preset delivery point is zero.

[0084] Furthermore, for each preset delivery point, the number of existing bicycles represents the bicycles that have been parked at the preset delivery point, and the number of bicycles to be delivered represents the number of bicycles required by the preset delivery point. In order to determine whether the bicycle data already parked at the preset delivery point can meet the number of bicycles required, a comparison is made between the number of existing bicycles and the number of bicycles to be delivered to determine the size relationship between the two. If the comparison determines that the size relationship is that the number of existing bicycles is greater than the number of bicycles to be delivered, it means that the number of bicycles at the preset delivery point exceeds the number of bicycles required, and the excess number of bicycles needs to be transferred out of the preset delivery point, so the preset delivery point is generated based on the excess number of bicycles. At the same time, in order to improve the scheduling efficiency of shared bicycles, the neighboring delivery points near the preset delivery point are searched for nearby scheduling. A preset range is set in advance, and the preset delivery point is taken as the center, and the preset delivery points within its preset range are searched as neighboring delivery points. Then, based on the information to be transferred out and the neighboring delivery points found, the scheduling information is generated to dispatch the bicycles that exceed the number to the neighboring delivery points according to the scheduling information.

[0085] Furthermore, if it is determined through comparison that the number of existing bicycles is less than the number of bicycles to be delivered, it means that the number of bicycles at the preset delivery point is less than the number of bicycles it needs, and it is necessary to transfer shared bicycles from other preset delivery points that exceed the number to the preset delivery point. Therefore, the information to be transferred is generated for the preset delivery point based on the number of missing bicycles. Similarly, in order to improve the scheduling efficiency of shared bicycles, the neighboring delivery points near the preset delivery point are searched for nearby scheduling. A preset range is set in advance, and the preset delivery point is taken as the center, and the preset delivery points within its preset range are searched as neighboring delivery points. Then, based on the information to be transferred and the neighboring delivery points found, the scheduling information is generated to transfer the missing number of bicycles from the neighboring delivery points according to the scheduling information.

[0086] Wherein, the step of generating the scheduling information according to the information to be called out and the adjacent delivery points corresponding to the preset delivery points includes:

[0087] Step S341, searching for a first-category adjacent delivery point for which scheduling information has not been generated from each of the adjacent delivery points, and searching for a second-category adjacent delivery point with information to be transferred from the first-category adjacent delivery points;

[0088] Step S342, determining the position distances between the preset delivery point and each of the second-category adjacent delivery points, and determining a target adjacent delivery point from each of the second-category adjacent delivery points according to each of the position distances;

[0089] Step S343: Generate the information to be called out and the target adjacent delivery point into the scheduling information.

[0090] Understandably, in the process of producing scheduling information, the adjacent delivery points being searched may have formed scheduling information with other preset delivery points. For example, the preset delivery point currently being processed is W1, and its adjacent delivery points after searching include W2, W3, and W4. At the same time, W2 and W3 are also adjacent delivery points of the preset delivery point W0. In the process of generating the scheduling information of W0, W2 has been determined as the target adjacent delivery point of W0, and is generated as scheduling information together with W0. Therefore, for the current preset delivery point W1, W2 has formed scheduling information with other preset delivery points. In order to avoid duplication, W2 is no longer considered as the adjacent delivery point of W1.

[0091] Furthermore, a scheduling identifier is added to the preset delivery point for which the scheduling information has been generated. For the preset delivery point currently being processed, the corresponding adjacent delivery points are identified, and the adjacent delivery points for which the scheduling identifier has not been added are determined, and this type of adjacent delivery points are determined as the first type of adjacent delivery points for which the scheduling information has not been generated. In addition, considering that the first type of adjacent delivery points may include adjacent delivery points whose existing number of shared bicycles does not meet their needs, it may also include adjacent delivery points whose existing number of shared bicycles exceeds their needs. For the preset delivery point currently being processed that contains information to be called out, it is necessary to call out shared bicycles, so it is necessary to find the adjacent delivery points from the first type of adjacent delivery points that have the number of shared bicycles that do not meet their needs and need to be called in. This type of adjacent delivery points carries information to be called in, so it is possible to find the second type of adjacent delivery points that carry information to be called in from the first type of adjacent delivery points by judging whether the first type of adjacent delivery points carry information to be called in.

[0092] Furthermore, considering that there are multiple second-category adjacent delivery points found, in order to shorten the scheduling distance, the position distance between the currently processed preset delivery point and each second-category adjacent delivery point is determined, and then a comparison is made between each position distance to determine the target adjacent delivery point with the shortest distance. Of course, if there is only one second-category adjacent delivery point found, the second adjacent delivery point found is directly used as the target adjacent delivery point. And, if there is no second-category adjacent delivery point, the preset range can be expanded to increase the adjacent delivery points for search. After obtaining the target adjacent delivery point, the to-be-called information of the currently processed preset delivery point and the target adjacent delivery point can be generated as scheduling information, so that the shared bicycles that exceed the number of preset delivery points can be transferred to the target adjacent delivery point according to the position of the target adjacent delivery point according to the scheduling information.

[0093] In addition, it should be noted that in the process of searching for the first type of adjacent delivery points that have not yet generated scheduling information, the adjacent delivery points that may have generated scheduling information may not fully meet their transfer-in quantity requirements. For example, for the above-mentioned adjacent delivery point W2, if the number of vehicles to be transferred out in the information to be transferred out of the preset delivery point W0 is 20, and the number of vehicles to be transferred in the information to be transferred in of W2 is 50, then even if W2 and W0 are generated as scheduling information together, W2 still has 30 vacancies. At this time, while adding a scheduling mark for this type of delivery point, its vacancy quantity can be marked, so that this type of delivery point can be used as the first type of adjacent delivery point again based on the vacancy quantity, avoiding duplication while ensuring that the delivery quantity requirements of each preset delivery point are met, and facilitating nearby scheduling to improve scheduling efficiency.

[0094] Understandably, for each preset delivery point, the number of bicycles to be delivered represents the number of shared bicycles required, and the number of existing bicycles represents the number of bicycles located at the preset delivery point. However, in the process of using shared bicycles, people may park shared bicycles from other preset delivery points at the preset delivery point, which makes the number of bicycles in the to-be-called-out information and the to-be-called-in information generated based on the size relationship between the number of bicycles to be delivered and the number of existing bicycles inaccurate. At the same time, factors such as weather and bicycle damage will also affect people's riding, thereby affecting the number of bicycles to be delivered, and further affecting the accuracy of the number of bicycles in the scheduling information. Therefore, in order to improve the accuracy of the number of bicycles to be dispatched in the generated scheduling information, this embodiment is provided with a correction mechanism for the scheduling information. Specifically, the step of generating the scheduling information of each preset delivery point based on the circulation data and the number of bicycles to be delivered includes:

[0095] Step S35, for each of the preset delivery points, obtaining the riding change parameters and optimization evaluation coefficients of the preset delivery point in multiple historical preset time periods, as well as the weather parameters in the next preset time period;

[0096] Step S36, searching for the historical damage ratio of shared bicycles in the preset area, and updating and calculating the quantity information in the scheduling information according to the historical damage ratio, riding change parameters, weather parameters and optimization evaluation coefficient, to obtain the updated scheduling information of the preset delivery point.

[0097] Furthermore, for each preset delivery point, its riding change parameters and optimization evaluation coefficients in multiple historical preset time periods are obtained, and its weather parameters in the next preset time period are also obtained. Among them, the riding change parameter represents the parameter of the actual riding volume change of the preset delivery point in each historical preset time period. For example, for the current preset time period t4, there are three historical preset time periods t1, t2, and t3. The actual riding volumes of t1, t2, t3, and t4 are r1, r2, r3, and r4 respectively. Then, the riding change parameter of t2 is the preset parameter corresponding to the difference (r2-r1), the riding change parameter of t3 is the preset parameter corresponding to the difference (r3-r2), and the riding change parameter of t4 is the preset parameter corresponding to the difference (r4-r3). For the optimization evaluation coefficient, it is a coefficient that reflects the difference between the number of corrections to the scheduling information in each previous historical preset time period and the actual number. For example, for the above t1, t2, t3, and t4, the number of modified dispatch information is R1, R2, R3, and R4 respectively, then the optimization evaluation coefficients of t2, t3, and t4 are the preset coefficients corresponding to (r1-R1), (r2-R2), (r3-R3), and (r4-R4) respectively. For weather parameters, it is a parameter that indicates the impact of good or bad weather on people's riding. For example, if it does not rain during the preset period and the temperature is suitable, the weather has little impact on people's riding; if it does not rain during the preset period but the weather is hot or cold, the weather has a greater impact on people's riding; if it rains heavily during the preset period, the weather has a serious impact on people's riding. The correspondence between various types of weather and parameters can be set in advance. After obtaining the weather, the corresponding weather parameters can be obtained by searching the corresponding relationship.

[0098] Furthermore, considering that shared bicycles may be damaged during use, the number of damages can be reflected by the previous damage ratio, so the historical damage ratio of shared bicycles in the preset area is searched. The historical damage ratio is obtained by statistically analyzing the proportion of damages in each historical period in the preset area. The overall historical damage ratio reflects the number of damages at each preset delivery point, reducing the inaccuracy of a single preset delivery point. The historical damage ratio is then combined with the riding change parameters, weather parameters and optimization evaluation coefficients to update the quantity information in the scheduling information. The specific calculation formula can be found in the following formula (1).

[0099]

[0100] Among them, yi represents the quantity information of the i-th preset delivery point, represents the weather parameters of the i-th preset delivery point, θ mi represents the optimization evaluation coefficient of the ith preset delivery point in the mth historical preset period, ε represents the historical damage ratio of the ith preset delivery point, τ miIt represents the riding change parameter of the i-th preset delivery point in the m-th historical preset time period, and n represents the number of historical preset time periods.

[0101] The quantity information in the dispatching information reflects the number of shared bicycles that need to be dispatched at the preset delivery point. By combining various factors that affect the demand for shared bicycles, the quantity information is updated and calculated, so that the quantity information more accurately reflects the number of shared bicycles that need to be dispatched at the preset delivery point, thereby performing more accurate dispatching.

[0102] This embodiment generates dispatch information reflecting the number of shared bicycles required to be dispatched for each preset delivery point based on the number of bicycles to be delivered and the number of bicycles already parked, and updates and calculates the number of shared bicycles required based on various factors affecting the number of shared bicycles required, thereby achieving accurate dispatch. At the same time, dispatching is performed nearby by searching for nearby delivery points, thereby improving dispatch efficiency and saving dispatch costs.

[0103] For further information, please refer to Figure 3 Based on the first and second embodiments of the public facility management and control method based on the Internet of Things of the present invention, a third embodiment of the public facility management and control method based on the Internet of Things of the present invention is proposed.

[0104] The third embodiment of the public facility management and control method based on the Internet of Things is different from the first and second embodiments of the public facility management and control method based on the Internet of Things in that the step of determining the maintenance information of the sports facility according to each of the facility images includes:

[0105] Step S41, for each of the facility images, obtaining three primary color values ​​of pixels in the facility image, and converting the three primary color values ​​of each of the pixels into grayscale values ​​to obtain a grayscale image corresponding to the facility image;

[0106] Step S42, extracting the contour image corresponding to the sports facility in the grayscale image according to the grayscale value of each pixel in the grayscale image, and obtaining the contour images of the sports facility at multiple viewing angles;

[0107] Step S43, transmitting each of the contour images to a preset recognition model, and processing a pixel matrix formed by the grayscale values ​​of each pixel in each of the contour images based on the first convolution kernel and the first pooling layer parameters in the first convolution layer of the preset recognition model to obtain a type matrix for determining the type of the item to which the sports facility belongs;

[0108] Step S44, based on the second convolution kernel and the second pooling layer parameters in the second convolution layer corresponding to the object type in the preset recognition model, the type matrix is ​​calculated to obtain a damage probability, and whether the sports facility is currently damaged is determined according to the damage probability;

[0109] Step S45: if there is damage, generating the maintenance information according to the damage type corresponding to the damage probability.

[0110] This embodiment obtains maintenance information of sports facilities by processing facility images. Specifically, in the process of collecting facility images of sports facilities, there may be moving people or objects passing through the location of the sports facilities, causing occlusion of the sports facilities, so that the collected facility images cannot reflect the status of the sports facilities, so it is necessary to eliminate such facility images. After elimination, for each perspective facility image of each sports facility, the three primary color values ​​of the pixels in each facility image are obtained respectively. Among them, the three primary colors refer to the three basic colors of red (R), green (G), and blue (B) that cannot be decomposed in color, and the three primary color values ​​are R value, G value and B value. Then, the three primary color values ​​of each pixel in each facility image are converted into grayscale values. The conversion can be performed according to the formula G=0.299*R+0.587*G+0.114*B. After the three primary color values ​​of each pixel in the facility image are converted into grayscale values, the facility map is correspondingly converted into a grayscale map.

[0111] Furthermore, in addition to the images of sports facilities, the grayscale image also contains other environmental factors, such as trees, people exercising, etc. In order to accurately obtain the appearance of the sports facilities, it is necessary to extract the outline of the sports facilities from the grayscale image. Different objects are imaged, and the grayscale values ​​of the contour pixels are different. The grayscale values ​​of each pixel in the grayscale image are identified and processed by a preset algorithm to determine the contour pixels of the sports facilities in the grayscale image, and then the contour image corresponding to the sports facilities in the grayscale image is extracted based on the contour pixels. After the grayscale values ​​of the pixels in the grayscale images corresponding to the facility images at each viewing angle are identified and processed, the contour images of the sports facilities at each viewing angle can be obtained to reflect the three-dimensional contour state of the sports facilities at each viewing angle.

[0112] Furthermore, a preset recognition model for recognizing contour images is provided in advance through training, and the preset recognition model includes a first convolution layer and a first pooling layer for recognizing the type of sports facilities, and a second convolution layer and a second pooling layer for recognizing whether the sports facilities are damaged. Each extracted contour image is transmitted to the preset recognition model, and the number of neurons in the input layer of the preset recognition model is consistent with the number of contour images. Each contour image is transmitted to a neuron, and the grayscale value of each pixel therein is formed into a pixel matrix by the neuron, and combined with the first convolution kernel of the first convolution layer and the first pooling layer parameters, the pixel matrix is ​​processed such as convolution operation and pooling operation to obtain a type matrix for determining the type of item to which the sports facilities belong. Specifically, the type of item to which the sports facilities belong can be determined by a full connection operation of the type matrix.

[0113] Furthermore, the preset recognition model includes multiple groups of second convolutional layers and second pooling layers, and different groups of second convolutional layers and second pooling layers are used to identify the damage of different types of items. After determining the type of item to which the currently processed sports facility belongs through the type matrix, the second convolutional layer and second pooling layer corresponding to the item type in the preset recognition model are searched, and the type matrix is ​​calculated through the second convolution kernel and second pooling layer parameters in the second convolutional layer to obtain the damage probability. Among them, the number of neurons in the output layer set by the preset recognition model is consistent with the number of damage types. After calculating the type matrix, each neuron in the output layer will output a probability value, indicating the probability of damage corresponding to the neuron in the sports facility. The various probability values ​​together constitute the calculated damage probability. The specific calculation formula can refer to the following formula (2).

[0114]

[0115] Where p(o) represents the damage probability value, k2 represents the second convolution kernel parameter, k1 represents the first convolution kernel parameter, and W i (i 1 , j 1 ) represents the pixel matrix of the i-th contour image. 1 row, j 1 The value of the column, S1(i 1 , j 1 ) represents the i-th convolution kernel in the first convolution kernel 1 row, j 1 The value of the column, α represents the step size corresponding to the first convolution kernel, m k1 represents the first pooling layer parameters,

[0116] Indicates the i-th 2 row, j 2 The value of the column, S2(i 2 , j 2 ) represents the i-th convolution kernel in the second convolution kernel 2 row, j 2 The value of the column, β represents the step size corresponding to the second convolution kernel, m k2 Represents the parameters of the second pooling layer. It should be noted that the first convolution kernel S1(i1,j1) and the second convolution kernel S2(i2,j2) are both in the form of matrices with equal numbers of rows and columns, and their parameters are the number of rows or columns, that is, if the first convolution kernel is a 3*3 matrix, then its first convolution kernel parameter k1 is equal to 3. The maximum values ​​of i1 and j1 are both k1, and the maximum values ​​of i2 and j2 are both k2.

[0117] Furthermore, the damage probability composed of various probability values ​​is used to indicate whether the sports facilities are damaged and the type of damage. Comparisons can be made between various probability values ​​to determine the maximum value. Then, the damage type corresponding to the neuron from which the maximum value comes is found, indicating that the sports facilities are most likely to have damage of the damage type. At the same time, a preset threshold representing the size of the possibility of damage is pre-set, and the maximum value is compared with the preset threshold to determine whether the maximum value is greater than the preset threshold. If it is greater than the preset threshold, it means that the sports facilities have this type of damage. In this way, it is determined that there is damage, and the searched damage type is used as the damage type corresponding to the damage probability, and is generated as maintenance information together with the geographical location of the sports facilities.

[0118] Furthermore, in order to enable the preset recognition model to directly recognize the grayscale image, the present embodiment is also provided with a mechanism for training the preset recognition model by using the grayscale image of the sports facility recognized as damaged as a training sample. Specifically, if there is damage, the step of generating the maintenance information according to the damage type corresponding to the damage probability includes:

[0119] Step S46, identifying the grayscale image of the damaged sports facilities as training data, and updating and training the preset recognition model according to the training data to obtain a training result;

[0120] Step S47, calculating the loss function value of the preset recognition model according to the training result;

[0121] Furthermore, a label is added to the grayscale image of the damaged sports facilities, and the content of the label at least includes the parameters of the minimum bounding rectangle of the sports facilities in the grayscale image. The minimum bounding rectangle substantially identifies the outline of the sports facilities in the grayscale image, and its parameters are equivalent to the outline parameters. The grayscale image with the label is then identified as training data, and the preset recognition model is updated and trained using the training data, and the minimum bounding rectangle of the sports facilities in the grayscale image is identified by the preset recognition model. In addition, a preset number of update trainings is pre-set, and when the number of update trainings reaches the preset number, the minimum bounding rectangle of the sports facilities in the grayscale image identified by the last training is used as the training result generated by the update training.

[0122] Furthermore, in order to reflect the training effect, a loss function is pre-set in the preset recognition model, and the training result is used as a parameter of the loss function. The loss function value is calculated, and the calculated loss function value reflects the effect of the preset recognition model on the grayscale image processing. The specific calculation formula can be found in the following formula (3).

[0123]

[0124] Where L represents the loss function, A represents the overlap area ratio between the first minimum bounding rectangle corresponding to the true outline of the sports facilities in the grayscale image and the second minimum bounding rectangle in the training result, and d z The center distance between the center point of the first minimum enclosing rectangle and the center point of the second minimum enclosing rectangle, d s It represents the distance between the farthest sides of the first minimum bounding rectangle and the second minimum bounding rectangle, that is, the distance between the farthest sides of the two rectangles, a represents the preset formula parameter, w1 and h1 represent the width and length of the first minimum bounding rectangle, and w2 and h2 represent the width and length of the second minimum bounding rectangle. It should be noted that the width direction corresponding to the width value and the length direction corresponding to the length value are the horizontal direction and the vertical direction of the pixel in the grayscale image.

[0125] Step S48, updating the grayscale image recognition weight matrix of the preset recognition model according to the size relationship between the loss function value and the preset threshold.

[0126] Furthermore, a preset threshold indicating the quality of the training effect is pre-set, and the calculated loss function value is compared with the preset threshold to determine the size relationship between the two, and based on the size relationship, the model parameters related to grayscale image recognition in the preset recognition model, that is, the grayscale image recognition weight matrix, are updated. Specifically, the step of updating the grayscale image recognition weight matrix of the preset recognition model based on the size relationship between the loss function value and the preset threshold includes:

[0127] Step S481, determining whether the magnitude relationship is that the loss function value is greater than a preset threshold,

[0128] Step S482, if the loss function value is greater than a preset threshold, the grayscale image recognition weight matrix of the preset recognition model is updated based on a preset update formula, and based on the updated grayscale image recognition weight matrix, the preset recognition model is updated and trained according to the training data to obtain a training result.

[0129] Furthermore, it is determined whether the magnitude relationship between the loss function value and the preset threshold is that the loss function value is greater than the preset threshold. If the loss function value is greater than the preset threshold, it means that the minimum bounding rectangle obtained by the preset recognition model for grayscale image recognition processing is greatly different from the minimum bounding rectangle in the label, and the recognition processing effect of the preset recognition model on the grayscale image is relatively poor. At this time, it is necessary to update the grayscale image recognition weight matrix related to grayscale image recognition in the preset recognition model. Among them, a preset update formula is pre-set, and the new grayscale image recognition weight matrix in the preset recognition model is obtained by calculating the preset update formula. The preset update formula can be specifically referred to in the following formula (4).

[0130]

[0131] Among them, V(i 3 , j 3 ) represents the grayscale image recognition weight matrix, x(i 3 , j 3 ) represents the i-th grayscale image recognition weight matrix 3 row, j 3 The value of the column, k3 represents the number of rows and columns of the grayscale image recognition weight matrix, Z(i 3 , j 3 ) represents the preset bias matrix.

[0132] Furthermore, according to the new grayscale image recognition weight matrix, the preset recognition model is updated and trained by the training data, and the minimum circumscribed matrix for the sports facilities in the grayscale image is obtained as the recognition result, which is used to calculate the loss function value and compare it with the preset threshold value until the loss function value is less than the preset threshold value. Once it is determined that the loss function value is less than the preset threshold value, it means that the preset recognition model recognizes the minimum circumscribed rectangle of the sports facilities in the grayscale image more accurately, which can be used for subsequent accurate processing of the grayscale image, so the update training of the preset recognition model is stopped.

[0133] This embodiment uses a preset recognition model to identify whether a sports facility is damaged and the type of damage according to the type of sports facility, so that the identification of the damage type is associated with the type of sports facility, and the accuracy of classification and recognition is higher. At the same time, the grayscale image identified as damaged is marked as training data to train the preset recognition model, so that the preset recognition model has the function of directly identifying and processing grayscale images. On the one hand, the collection of training samples is more convenient and accurate, and on the other hand, the training time can be shortened, making the function of the preset recognition model richer.

[0134] In addition, an embodiment of the present invention further provides a public facility management and control system based on the Internet of Things, and the public facility management and control system based on the Internet of Things includes an Internet of Things server. Figure 4 , Figure 4 It is a structural schematic diagram of the equipment hardware operating environment involved in the implementation scheme of the public facility management and control system based on the Internet of Things of the present invention.

[0135] like Figure 4As shown, the Internet of Things server in the public facility management and control system based on the Internet of Things may include: a processor 1001, such as a CPU, a communication bus 1002, a user interface 1003, a network interface 1004, and a storage 1005. Among them, the communication bus 1002 is used to realize the connection and communication between these components. The user interface 1003 may include a display screen (Display), an input unit such as a keyboard (Keyboard), and the optional user interface 1003 may also include a standard wired interface and a wireless interface. The network interface 1004 may optionally include a standard wired interface and a wireless interface (such as a WI-FI interface). The storage 1005 may be a high-speed RAM storage, or it may be a stable storage (non-volatile memory), such as a disk storage. The storage 1005 may also be a storage device independent of the aforementioned processor 1001.

[0136] Those skilled in the art will understand that Figure 4 The hardware structure of the Internet of Things server in the public utility management and control system based on the Internet of Things shown in the figure does not constitute a limitation on the public utility management and control system based on the Internet of Things, and may include more or fewer components than shown in the figure, or a combination of certain components, or a different arrangement of components.

[0137] like Figure 4 As shown, the storage 1005 as a storage medium may include an operating system, a network communication module, a user interface module and a control program. The operating system is a program for managing and controlling the IoT server and software resources in the public facility management and control system based on the IoT, and supports the operation of the network communication module, the user interface module, the control program and other programs or software; the network communication module is used to manage and control the network interface 1004; the user interface module is used to manage and control the user interface 1003.

[0138] exist Figure 4 In the hardware structure of the Internet of Things server of the public facility management and control system based on the Internet of Things shown, the network interface 1004 is mainly used to connect to other system servers and communicate data with other system servers; the user interface 1003 is mainly used to connect to the client (user end) and communicate data with the client; the processor 1001 can call the control program stored in the storage 1005 and perform the following operations:

[0139] When the IoT server detects a control instruction triggered based on a preset control cycle, it identifies the type identifier carried by the control instruction, and determines the type of the public facility to be controlled according to the type identifier, where different types of public facilities correspond to different preset control cycles;

[0140] If the type of the public facility to be controlled is a shared bicycle type, then the flow data of the shared bicycles in the preset area within the preset time period is obtained, and according to the flow data, the number of shared bicycles to be placed in each preset placement point in the preset area is determined;

[0141] Generate scheduling information for each preset delivery point according to the flow data and the number of bicycles to be delivered, and schedule the shared bicycles in the preset area according to the scheduling information to manage and control the shared bicycles in the preset area;

[0142] If the type of the public facility to be controlled is a sports facility, facility images of the sports facilities in a preset area from multiple perspectives are obtained, and maintenance information of the sports facilities is determined based on each of the facility images, and the sports facilities in the preset area are controlled based on the maintenance information.

[0143] Furthermore, the step of generating the scheduling information of each preset delivery point according to the flow data and the number of bicycles to be delivered includes:

[0144] Determine the final destination of the shared bicycle within the preset time period according to the identification time of the location identification carried by the circulation data, and count the number of shared bicycles at each final destination to obtain the number of existing bicycles at each preset delivery point;

[0145] For each of the preset delivery points, determining the relationship between the number of existing bicycles and the number of bicycles to be delivered;

[0146] If the size relationship is that the number of existing bicycles is greater than the number of bicycles to be delivered, generating information to be called out for the preset delivery point, and generating the scheduling information according to the information to be called out and the adjacent delivery points corresponding to the preset delivery point;

[0147] If the size relationship is that the number of existing bicycles is less than the number of bicycles to be deployed, the pending transfer information is generated for the preset delivery point, and the scheduling information is generated based on the pending transfer information and the adjacent delivery points corresponding to the preset delivery point.

[0148] Furthermore, the step of generating the scheduling information according to the information to be called out and the adjacent delivery points corresponding to the preset delivery points includes:

[0149] Find out the first-category adjacent delivery points that have not generated scheduling information from each of the adjacent delivery points, and find out the second-category adjacent delivery points that carry information to be transferred from the first-category adjacent delivery points;

[0150] Determine the position distances between the preset delivery point and each of the second-category adjacent delivery points, and determine a target adjacent delivery point from each of the second-category adjacent delivery points according to each of the position distances;

[0151] The information to be called out and the target adjacent delivery point are generated as the scheduling information.

[0152] Furthermore, after the step of generating the scheduling information of each preset delivery point according to the flow data and the number of bicycles to be delivered, the following steps are included:

[0153] For each of the preset delivery points, obtaining the riding change parameters and optimization evaluation coefficients of the preset delivery point in multiple historical preset time periods, as well as the weather parameters in the next preset time period;

[0154] Find the historical damage ratio of shared bicycles in the preset area, and update the quantity information in the scheduling information according to the historical damage ratio, riding change parameters, weather parameters and optimization evaluation coefficient to obtain the updated scheduling information of the preset delivery point, wherein the update calculation formula is:

[0155]

[0156] Among them, yi represents the quantity information of the i-th preset delivery point, represents the weather parameters of the i-th preset delivery point, θ mi represents the optimization evaluation coefficient of the ith preset delivery point in the mth historical preset period, ε represents the historical damage ratio of the ith preset delivery point, τ mi It represents the riding change parameter of the i-th preset delivery point in the m-th historical preset time period, and n represents the number of historical preset time periods.

[0157] Furthermore, the step of determining the number of shared bicycles to be placed in each preset placement point in the preset area according to the flow data includes:

[0158] Identify the location identifiers carried by each item of the flow data, and count the number of identical location identifiers to obtain multiple statistical results;

[0159] According to the correspondence between the position identifier of each statistical result and each preset delivery point, the number of shared bicycles to be delivered at each preset delivery point is determined.

[0160] Furthermore, the step of determining the maintenance information of the sports facilities according to each of the facility images includes:

[0161] For each of the facility images, obtaining three primary color values ​​of pixels in the facility image, and converting the three primary color values ​​of each of the pixels into grayscale values ​​to obtain a grayscale image corresponding to the facility image;

[0162] Extracting a contour image corresponding to the sports facility in the grayscale image according to the grayscale value of each pixel in the grayscale image, and obtaining contour images of the sports facility at multiple viewing angles;

[0163] Transmitting each of the contour images to a preset recognition model, and processing a pixel matrix formed by the grayscale values ​​of each pixel in each of the contour images based on the first convolution kernel and the first pooling layer parameters in the first convolution layer of the preset recognition model to obtain a type matrix for determining the type of the item to which the sports facility belongs;

[0164] Based on the second convolution kernel and the second pooling layer parameters in the second convolution layer corresponding to the item type in the preset recognition model, the type matrix is ​​calculated to obtain the damage probability, and it is judged whether the current sports facility is damaged according to the damage probability, wherein the calculation formula is:

[0165]

[0166] Where p(o) represents the damage probability value, k2 represents the second convolution kernel parameter, k1 represents the first convolution kernel parameter, and W i (i 1 , j 1 ) represents the pixel matrix of the i-th contour image. 1 row, j 1 The value of the column, S1(i 1 , j 1 ) represents the i-th convolution kernel in the first convolution kernel 1 row, j 1 The value of the column, α represents the 0 step length corresponding to the first convolution kernel, m k1 represents the first pooling layer parameters, Indicates the i-th 2 row, j 2 The value of the column, S2(i 2 , j 2 ) represents the i-th convolution kernel in the second convolution kernel 2 row, j 2 The value of the column, β represents the step size corresponding to the second convolution kernel, m k2 Represents the parameters of the second pooling layer.

[0167] If there is damage, the repair information is generated according to the damage type corresponding to the damage probability.

[0168] Furthermore, if there is damage, the step of generating the repair information according to the damage type corresponding to the damage probability includes:

[0169] Identify the grayscale image of the damaged sports facilities as training data, and update and train the preset recognition model according to the training data to obtain a training result;

[0170] The loss function value of the preset recognition model is calculated according to the training result, and the calculation formula is:

[0171]

[0172] Where L represents the loss function, A represents the overlap area ratio between the first minimum bounding rectangle corresponding to the true outline of the sports facilities in the grayscale image and the second minimum bounding rectangle in the training result, and d z The center distance between the center point of the first minimum enclosing rectangle and the center point of the second minimum enclosing rectangle, d s represents the distance between the farthest edges of the first minimum circumscribed rectangle and the second minimum circumscribed rectangle, a represents a preset formula parameter, w1 and h1 represent the width and length of the first minimum circumscribed rectangle, and w2 and h2 represent the width and length of the second minimum circumscribed rectangle;

[0173] According to the magnitude relationship between the loss function value and the preset threshold, the grayscale image recognition weight matrix of the preset recognition model is updated.

[0174] Furthermore, the step of updating the grayscale image recognition weight matrix of the preset recognition model according to the magnitude relationship between the loss function value and the preset threshold value includes:

[0175] Determine whether the magnitude relationship is that the loss function value is greater than a preset threshold;

[0176] If the loss function value is greater than a preset threshold, the grayscale image recognition weight matrix of the preset recognition model is updated based on a preset update formula, so as to perform the step of updating and training the preset recognition model according to the training data based on the updated grayscale image recognition weight matrix to obtain a training result;

[0177] Among them, the preset update formula is:

[0178]

[0179] Among them, V(i 3 , j 3 ) represents the grayscale image recognition weight matrix, x(i 3 , j 3 ) represents the i-th grayscale image recognition weight matrix3 row, j 3 The value of the column, k3 represents the number of rows and columns of the grayscale image recognition weight matrix, Z(i 3 , j 3 ) represents the preset bias matrix.

[0180] The specific implementation methods of the public facility management and control system based on the Internet of Things of the present invention are basically the same as the above-mentioned embodiments of the public facility management and control method based on the Internet of Things, and will not be repeated here.

[0181] The embodiment of the present invention further provides a storage medium having a control program stored thereon, and when the control program is executed by a processor, the steps of the public facility control method based on the Internet of Things are implemented as described above.

[0182] The storage medium of the present invention may be a computer-readable storage medium, and its implementation method is basically the same as the above-mentioned embodiments of the public facility management and control method based on the Internet of Things, and will not be repeated here.

[0183] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the enlightenment of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the present invention and the claims. All equivalent structures or equivalent process changes made using the contents of the specification and drawings of the present invention, or directly or indirectly used in other related technical fields, are protected by the present invention.

Claims

1. A method for public facility management and control based on the Internet of Things, characterized in that: Applied to an Internet of Things server, the public facility management and control method includes: When the IoT server detects a control instruction triggered based on a preset control cycle, it identifies the type identifier carried by the control instruction, and determines the type of the public facility to be controlled according to the type identifier, where different types of public facilities correspond to different preset control cycles; If the type of the public facility to be controlled is a shared bicycle type, then the flow data of the shared bicycles in the preset area within the preset time period is obtained, and according to the flow data, the number of shared bicycles to be placed in each preset placement point in the preset area is determined; Generate scheduling information for each preset delivery point according to the flow data and the number of bicycles to be delivered, and schedule the shared bicycles in the preset area according to the scheduling information to manage and control the shared bicycles in the preset area; If the type of the public facility to be controlled is a sports facility, facility images of the sports facilities in a preset area from multiple perspectives are obtained, and maintenance information of the sports facilities is determined based on each of the facility images, and the sports facilities in the preset area are controlled based on the maintenance information.

2. The public facility management and control method according to claim 1, characterized in that: The step of generating the scheduling information of each preset delivery point according to the flow data and the number of bicycles to be delivered comprises: Determine the final destination of the shared bicycle within the preset time period according to the identification time of the location identification carried by the circulation data, and count the number of shared bicycles at each final destination to obtain the number of existing bicycles at each preset delivery point; For each of the preset delivery points, determining the relationship between the number of existing bicycles and the number of bicycles to be delivered; If the size relationship is that the number of existing bicycles is greater than the number of bicycles to be delivered, generating information to be called out for the preset delivery point, and generating the scheduling information according to the information to be called out and the adjacent delivery points corresponding to the preset delivery point; If the size relationship is that the number of existing bicycles is less than the number of bicycles to be deployed, the pending transfer information is generated for the preset delivery point, and the scheduling information is generated based on the pending transfer information and the adjacent delivery points corresponding to the preset delivery point.

3. The public facility management and control method according to claim 2, characterized in that: The step of generating the scheduling information according to the information to be called out and the adjacent delivery points corresponding to the preset delivery points includes: Find out the first-category adjacent delivery points that have not generated scheduling information from each of the adjacent delivery points, and find out the second-category adjacent delivery points that carry information to be transferred from the first-category adjacent delivery points; Determine the position distances between the preset delivery point and each of the second-category adjacent delivery points, and determine a target adjacent delivery point from each of the second-category adjacent delivery points according to each of the position distances; The information to be called out and the target adjacent delivery point are generated as the scheduling information.

4. The public facility management and control method according to claim 1, characterized in that: After the step of generating the scheduling information of each preset delivery point according to the flow data and the number of bicycles to be delivered, the following steps are included: For each of the preset delivery points, obtaining the riding change parameters and optimization evaluation coefficients of the preset delivery point in multiple historical preset time periods, as well as the weather parameters in the next preset time period; Find the historical damage ratio of shared bicycles in the preset area, and update the quantity information in the scheduling information according to the historical damage ratio, riding change parameters, weather parameters and optimization evaluation coefficient to obtain the updated scheduling information of the preset delivery point, wherein the update calculation formula is: Among them, yi represents the quantity information of the i-th preset delivery point, represents the weather parameters of the i-th preset delivery point, θ mi represents the optimization evaluation coefficient of the ith preset delivery point in the mth historical preset period, ε represents the historical damage ratio of the ith preset delivery point, τ mi It represents the riding change parameter of the i-th preset delivery point in the m-th historical preset time period, and n represents the number of historical preset time periods.

5. The public facility management and control method according to claim 1, characterized in that: The step of determining the number of shared bicycles to be placed at each preset placement point in the preset area according to the flow data includes: Identify the location identifiers carried by each item of the flow data, and count the number of identical location identifiers to obtain multiple statistical results; According to the correspondence between the position identifier of each statistical result and each preset delivery point, the number of shared bicycles to be delivered at each preset delivery point is determined.

6. The public facility management and control method according to any one of claims 1 to 5, characterized in that: The step of determining the maintenance information of the sports facilities according to each of the facility images comprises: For each of the facility images, obtaining three primary color values ​​of pixels in the facility image, and converting the three primary color values ​​of each of the pixels into grayscale values ​​to obtain a grayscale image corresponding to the facility image; Extracting a contour image corresponding to the sports facility in the grayscale image according to the grayscale value of each pixel in the grayscale image, and obtaining contour images of the sports facility at multiple viewing angles; Transmitting each of the contour images to a preset recognition model, and processing a pixel matrix formed by the grayscale values ​​of each pixel in each of the contour images based on the first convolution kernel and the first pooling layer parameters in the first convolution layer of the preset recognition model to obtain a type matrix for determining the type of the item to which the sports facility belongs; Based on the second convolution kernel and the second pooling layer parameters in the second convolution layer corresponding to the item type in the preset recognition model, the type matrix is ​​calculated to obtain the damage probability, and it is judged whether the current sports facility is damaged according to the damage probability, wherein the calculation formula is: Where p(o) represents the damage probability value, k2 represents the second convolution kernel parameter, k1 represents the first convolution kernel parameter, and W i (i1, j1) represents the value of the i1th row and j1th column in the pixel matrix formed by the i-th contour image, S1(i1, j1) represents the value of the i1th row and j1th column in the first convolution kernel, α represents the step size corresponding to the first convolution kernel, m k1 represents the first pooling layer parameters, represents the value of the i2th row and j2th column in the type matrix, S2(i2, j2) represents the value of the i2th row and j2th column in the second convolution kernel, β represents the step size corresponding to the second convolution kernel, m k2 Represents the parameters of the second pooling layer. If there is damage, the repair information is generated according to the damage type corresponding to the damage probability.

7. The public facility management and control method according to claim 6, characterized in that: If damage exists, the step of generating the repair information according to the damage type corresponding to the damage probability includes: Identify the grayscale image of the damaged sports facilities as training data, and update and train the preset recognition model according to the training data to obtain a training result; The loss function value of the preset recognition model is calculated according to the training result, and the calculation formula is: Where L represents the loss function, A represents the overlap area ratio between the first minimum bounding rectangle corresponding to the true outline of the sports facilities in the grayscale image and the second minimum bounding rectangle in the training result, and d z The center distance between the center point of the first minimum enclosing rectangle and the center point of the second minimum enclosing rectangle, d s represents the distance between the farthest edges of the first minimum circumscribed rectangle and the second minimum circumscribed rectangle, a represents a preset formula parameter, w1 and h1 represent the width and length of the first minimum circumscribed rectangle, and w2 and h2 represent the width and length of the second minimum circumscribed rectangle; According to the magnitude relationship between the loss function value and the preset threshold, the grayscale image recognition weight matrix of the preset recognition model is updated.

8. The public facility management and control method according to claim 7, characterized in that: The step of updating the grayscale image recognition weight matrix of the preset recognition model according to the magnitude relationship between the loss function value and the preset threshold value comprises: Determine whether the magnitude relationship is that the loss function value is greater than a preset threshold; If the loss function value is greater than a preset threshold, the grayscale image recognition weight matrix of the preset recognition model is updated based on a preset update formula, so as to perform the step of updating and training the preset recognition model according to the training data based on the updated grayscale image recognition weight matrix to obtain a training result; Among them, the preset update formula is: Among them, V(i3,j3) represents the grayscale image recognition weight matrix, x(i3,j3) represents the value of the i3th row and j3th column in the grayscale image recognition weight matrix, k3 represents the number of rows and columns of the grayscale image recognition weight matrix, and Z(i3,j3) represents the preset bias matrix.

9. A public facility management and control system based on the Internet of Things, characterized in that: The public facility management and control system based on the Internet of Things includes an Internet of Things server, in which a storage, a processor, a communication bus and a control program stored on the storage are arranged: The communication bus is used to realize the connection and communication between the processor and the storage; The processor is used to execute the control program to implement the steps of the public facility management and control method based on the Internet of Things as described in any one of claims 1-8.

10. A storage medium, characterized in that: The storage medium stores a control program, and when the control program is executed by the processor, the steps of the public facility management and control method based on the Internet of Things as described in any one of claims 1 to 8 are implemented.

Images