Cloud printing method and terminal based on AP positioning

Abstract

The invention discloses a cloud printing method and terminal based on AP positioning, and the method comprises the steps: obtaining a non-feedback point after the position of a printer is updated in response to a position updating instruction of the printer; searching source fingerprint data of feedback points in a preset range of the non-feedback points; constructing an estimated fingerprint curved surface according to the non-feedback point and the source fingerprint data based on a radial basis function; determining an estimated AP signal value of an area corresponding to the non-feedback point according to the estimated fingerprint curved surface; inputting the estimated AP signal value into a pre-trained deep learning model to obtain estimated position information; generating standard fingerprint data according to the estimated AP signal value and the estimated position information; updating an initial fingerprint database according to the standard fingerprint data to obtain a standard fingerprint database; and in response to a printing instruction of a user, a target printer is positioned according to the standard fingerprint database and the printing instruction to complete printing operation, so that the workload of updating the fingerprint database is reduced, and the positioning precision is improved.

Description

Technical Field

[0001] The present invention relates to the field of cloud printing technology, and in particular to a cloud printing method and terminal based on AP positioning. Background Art

[0002] As cloud printing technology continues to improve, the demand for cloud printer matching is also increasing. Quickly, conveniently and accurately selecting nearby printers for printing is the primary requirement for customer experience. In cloud printing technology, if users find printers of different brands in the cloud printing environment, they cannot accurately and quickly locate the target printer when choosing to print.

[0003] In order to facilitate users to quickly find the nearest available printer, the strength of the wireless signal can be measured at the AP (Access Point) access point or wireless network card in the wireless local area network, and the printer can be located by matching the signal strength. Among them, the location fingerprint method is a technology for collecting wireless signals for indoor positioning. The location fingerprint method can collect signal strength based on RSSI (Received Signal Strength Indication) technology to build a fingerprint database for positioning. Specifically, the location information and signal strength of several AP fixed points set in the system coverage area are first collected offline, and the location information and signal strength are transmitted to the data center through a wired network to form a location fingerprint database, and then the real-time object to be tested is measured to obtain the signal strength, and finally the nearest neighbor method is used to analyze and match the position of the object to be tested. However, this positioning method is easily affected by environmental factors, and if the printer location information changes, the information needs to be re-collected and calibrated, the workload of collecting data is large, and if only part of the signal value is updated, the error will increase with the increase in the number of updated signal values, thereby affecting the positioning accuracy. Summary of the invention

[0004] The technical problem to be solved by the present invention is: a cloud printing method and terminal based on AP positioning, which can pre-update the fingerprint database when the printer position changes, thereby reducing the collection workload and ensuring the positioning accuracy.

[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0006] A cloud printing method based on AP positioning, comprising:

[0007] In response to the printer's position update instruction, obtaining the non-feedback point after the printer position is updated;

[0008] Searching for source fingerprint data of a feedback point located within a preset range of the non-feedback point;

[0009] Based on the radial basis function, constructing an estimated fingerprint surface according to the non-feedback points and the source fingerprint data;

[0010] Determining an estimated AP signal value of an area corresponding to the non-feedback point according to the estimated fingerprint surface;

[0011] Inputting the estimated AP signal value into a pre-trained deep learning model to obtain estimated location information;

[0012] Generate standard fingerprint data according to the estimated AP signal value and the estimated location information;

[0013] Update the initial fingerprint database according to the standard fingerprint data to obtain a standard fingerprint database;

[0014] In response to a user's printing instruction, a target printer is located according to the standard fingerprint database and the printing instruction to complete a printing operation.

[0015] In order to solve the above technical problems, another technical solution adopted by the present invention is:

[0016] A cloud printing terminal based on AP positioning includes a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, each step in the above-mentioned cloud printing method based on AP positioning is implemented.

[0017] The beneficial effect of the present invention is that when the position of the printer changes, the reference points in the current environment that need to update the fingerprint data are confirmed as non-feedback points, and the reference points that do not need to be updated are confirmed as feedback points. An estimated fingerprint surface is constructed based on the source fingerprint data of the feedback points near the non-feedback points, and the estimated AP signal values ​​around the non-feedback points are determined based on the estimated fingerprint surface. Then, through the pre-trained deep learning model, the estimated position information is inferred from the estimated AP signal value, so as to generate standard fingerprint data based on the estimated AP signal value and the estimated position information. In this way, the fingerprint data of the non-feedback points is adaptively updated without the need to re-collect large quantities of data, reducing time and labor cost overhead. At the same time, the initial fingerprint database is updated based on the standard fingerprint data of the non-feedback points to avoid the increase in errors caused by updating only some feedback points, thereby ensuring the positioning accuracy of the subsequent target printer. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 A flowchart of a cloud printing method based on AP positioning provided by an embodiment of the present invention;

[0019] Figure 2 A schematic diagram of the structure of a cloud printing terminal based on AP positioning provided by an embodiment of the present invention;

[0020] Description of labels:

[0021] 100. A cloud printing terminal based on AP positioning; 101. Memory; 102. Processor. DETAILED DESCRIPTION

[0022] In order to explain the technical content, achieved objectives and effects of the present invention in detail, the following is an explanation in combination with the implementation modes and the accompanying drawings.

[0023] An embodiment of the present invention provides a cloud printing method based on AP positioning, including:

[0024] In response to the printer's position update instruction, obtaining the non-feedback point after the printer position is updated;

[0025] Searching for source fingerprint data of a feedback point located within a preset range of the non-feedback point;

[0026] Based on the radial basis function, constructing an estimated fingerprint surface according to the non-feedback points and the source fingerprint data;

[0027] Determining an estimated AP signal value of an area corresponding to the non-feedback point according to the estimated fingerprint surface;

[0028] Inputting the estimated AP signal value into a pre-trained deep learning model to obtain estimated location information;

[0029] Generate standard fingerprint data according to the estimated AP signal value and the estimated location information;

[0030] Update the initial fingerprint database according to the standard fingerprint data to obtain a standard fingerprint database;

[0031] In response to a user's printing instruction, a target printer is located according to the standard fingerprint database and the printing instruction to complete a printing operation.

[0032] From the above description, it can be seen that the beneficial effect of the present invention is that when the position of the printer changes, the reference point in the current environment that needs to update the fingerprint data is confirmed as a non-feedback point, and the reference point that does not need to update the fingerprint data is a feedback point. An estimated fingerprint surface is constructed based on the source fingerprint data of the feedback point near the non-feedback point, and the estimated AP signal value around the non-feedback point is determined based on the estimated fingerprint surface, and then the estimated position information is inferred from the estimated AP signal value through the pre-trained deep learning model, so as to generate standard fingerprint data based on the estimated AP signal value and the estimated position information. In this way, the fingerprint data of the non-feedback point is adaptively updated without the need to re-collect large quantities of data, reducing time and labor cost overhead. At the same time, the initial fingerprint database is updated based on the standard fingerprint data of the non-feedback point to avoid the increase in errors caused by updating only some feedback points, thereby ensuring the positioning accuracy of the subsequent target printer.

[0033] Furthermore, before responding to the printer's location update instruction, the method further includes:

[0034] Determine the reference point for the initial position of the printer in the target environment;

[0035] Acquire the AP signal value and location information collected by the reference point;

[0036] An initial fingerprint database is constructed according to the reference point, the AP signal value and the location information.

[0037] From the above description, it can be seen that by recording the initial position of the printer and the corresponding AP signal value, an initial fingerprint database is constructed according to the initial position and the AP signal value. This positioning method can have a certain adaptability to environmental changes and avoid repeated data collection.

[0038] Further, the searching for source fingerprint data of feedback points located within the preset range of non-feedback points includes:

[0039] In the initial fingerprint database, a target reference point whose relative position relationship with the non-feedback point meets a preset range is searched, and the AP signal value and the position information of the target reference point are marked as source fingerprint data of the feedback point.

[0040] From the above description, it can be seen that since the position information of the non-feedback point is unknown, but the position change of the printer is generally small, the source fingerprint data of the reference point near the non-feedback point can be determined based on the relative position relationship between the non-feedback point and the reference point. There is no need to collect and calibrate the position information of the non-feedback point, thereby reducing the update workload.

[0041] Furthermore, the constructing of the estimated fingerprint surface based on the radial basis function according to the non-feedback points and the source fingerprint data includes:

[0042] Define a cut-off area as a circular area with the non-feedback point as the center and including the feedback point, and calculate the distance between the non-feedback point and each of the feedback points in the cut-off area;

[0043] Calculate the radial basis function value of each distance by using the radial basis function;

[0044] The cut-off area is grid-divided based on a preset interval distance, and each grid point is interpolated according to the radial basis function value and the source fingerprint data to obtain an estimated fingerprint surface.

[0045] From the above description, we can know that the radial basis function is used to interpolate the circular area with the non-feedback point as the center and including the feedback point, so as to obtain the estimated AP signal value of all grid points in the circular area, that is, the estimated fingerprint surface. In this way, the AP signal values ​​of the area near the non-feedback point are filled in batches, and a large amount of data is collected again, which reduces time and labor cost.

[0046] Furthermore, before inputting the estimated AP signal value into the pre-trained deep learning model to generate standard fingerprint data, the method further includes:

[0047] The estimated AP signal value is optimized and screened according to the cutoff region corresponding to each of the non-feedback points.

[0048] From the above description, it can be seen that the estimated AP signal value is optimized and screened based on the cutoff region to ensure the estimation accuracy.

[0049] Further, the optimizing and screening the estimated AP signal value according to the cut-off area corresponding to each of the non-feedback points includes:

[0050] Determine an overlapping area between the cut-off areas corresponding to each of the non-feedback points;

[0051] Gridding the overlapping area based on the preset interval distance, and determining the coordinate position of the estimated AP signal value according to the grid points;

[0052] The estimated AP signal value whose coordinate position is within the overlapping area is selected.

[0053] From the above description, it can be seen that since the estimated AP signal value of the non-feedback point is obtained based on the interpolation of the source fingerprint data in the cutoff zone, the estimation accuracy of the non-feedback point is related to the radius length of the cutoff zone and the area of ​​the overlapping area. Therefore, the estimated AP signal value that does not meet the radius length of the cutoff zone and the area of ​​the overlapping area is excluded to ensure the accuracy of the estimated signal value.

[0054] Furthermore, before inputting the estimated AP signal value into the pre-trained deep learning model to generate standard fingerprint data, the method further includes:

[0055] The estimated AP signal value is optimized and screened according to the source fingerprint data.

[0056] From the above description, it can be seen that the estimated AP signal value is optimized and screened based on the accuracy of the source fingerprint data to ensure the estimation accuracy.

[0057] Further, the source fingerprint data includes a timestamp;

[0058] The optimizing and screening the estimated AP signal value according to the source fingerprint data includes:

[0059] Determining a first priority of the source fingerprint data according to a sequence of the timestamps;

[0060] Determining, according to the first priority, a second priority of the estimated AP signal value obtained by interpolation based on the source fingerprint data;

[0061] The estimated AP signal value with the second priority as the highest level is selected.

[0062] From the above description, it can be seen that due to the change of time, the actual fingerprint data may vary greatly from the source fingerprint data, resulting in a decrease in the accuracy of the source fingerprint data. Therefore, for source fingerprint data with a relatively recent timestamp, the first priority of its source fingerprint data is marked as high to indicate that the accuracy of the source fingerprint data is relatively high. Since the estimated AP signal value is obtained by interpolation of the source fingerprint data, the accuracy of the estimated AP signal value is affected by the source fingerprint data, so the estimated AP signal value obtained by interpolation of the source fingerprint data with higher accuracy also has a higher corresponding accuracy. Therefore, the accuracy of the estimated AP signal value is ensured by excluding the source fingerprint data with low accuracy.

[0063] Further, locating the target printer according to the standard fingerprint database and the printing instruction includes:

[0064] Acquire a real-time AP signal value of a mobile terminal corresponding to the printing instruction;

[0065] Determine the scene type of the real-time AP signal value based on the scene analysis model, and obtain first candidate fingerprint data from the standard fingerprint database according to the scene type;

[0066] Determine, from the first candidate fingerprint data, second candidate fingerprint data closest to the real-time AP signal value by using a neighboring information model;

[0067] generating a candidate printer list according to the second candidate fingerprint data, and returning the list to the mobile terminal according to the printing instruction;

[0068] In response to a selection instruction of the mobile terminal, a target printer is determined in the candidate printer list.

[0069] From the above description, it can be seen that since the printer is located in many environmental locations, in order to improve the positioning efficiency, the search and matching scope is first narrowed down through the scene analysis model, and then the similar fingerprint data is matched through the proximity information model to ensure the accuracy of positioning.

[0070] Another embodiment of the present invention provides a cloud printing terminal based on AP positioning, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements each step of the above-mentioned cloud printing method based on AP positioning.

[0071] From the above description, it can be seen that the beneficial effect of the present invention is that when the position of the printer changes, the reference point in the current environment that needs to update the fingerprint data is confirmed as a non-feedback point, and the reference point that does not need to update the fingerprint data is a feedback point. An estimated fingerprint surface is constructed based on the source fingerprint data of the feedback point near the non-feedback point, and the estimated AP signal value around the non-feedback point is determined based on the estimated fingerprint surface, and then the estimated position information is inferred from the estimated AP signal value through the pre-trained deep learning model, so as to generate standard fingerprint data based on the estimated AP signal value and the estimated position information. In this way, the fingerprint data of the non-feedback point is adaptively updated without the need to re-collect large quantities of data, reducing time and labor cost overhead. At the same time, the initial fingerprint database is updated based on the standard fingerprint data of the non-feedback point to avoid the increase in errors caused by updating only some feedback points, thereby ensuring the positioning accuracy of the subsequent target printer.

[0072] The embodiment of the present invention provides a cloud printing method and terminal based on AP positioning, which can pre-update the fingerprint database when the printer position changes, reduce the collection workload and ensure the positioning accuracy. The following is an explanation through a specific embodiment:

[0073] Please refer to Figure 1 , Embodiment 1 of the present invention is:

[0074] A cloud printing method based on AP positioning, comprising:

[0075] S1. In response to a position update instruction of a printer, a non-feedback point after the printer position is updated is obtained.

[0076] In some embodiments, reference points in the current environment that need to be updated are identified as non-feedback points.

[0077] Specifically, before step S1, the method further includes:

[0078] S101, determining a reference point of an initial position of a printer in a target environment.

[0079] S102: Acquire the AP signal value and location information collected by the reference point.

[0080] S103: construct an initial fingerprint database according to the reference point, the AP signal value and the location information.

[0081] In some embodiments, standard coordinates are constructed in the target environment, and several points are evenly selected as reference points. AP signal value sampling is performed at these reference points to confirm the RSSI (signal strength) value received by the reference point. The RSSI value distribution characteristics of the reference point and the coordinate information of the reference point are stored together in a preset fingerprint database to obtain an initial fingerprint database.

[0082] In some embodiments, while entering the fingerprint database, the distribution location and number of AP devices in the target environment are determined. The RSSI ranging method used in AP positioning is a commonly used ranging method in wireless positioning networks. When the wireless signal propagates in the air, due to interference from various environmental or other factors, the RSSI value will also decay as the distance increases. Therefore, there is an error X0 in the RSSI value. By determining the distribution characteristics of AP devices in the target environment, the error X0 can be made close to 0. Specifically, Fingerprint data collected by optimizing AP positioning, where PL(d) represents the path loss after distance d, PL(d0) represents the path loss after unit distance d0, d0 is usually 1m, n represents the signal attenuation factor, which ranges from 2 to 4, and N0 represents a random number with a mean of 0 and a Gaussian distribution, with a standard deviation range of 4 to 10. The error X0 can be reduced through measurement and collection of multiple AP devices to accurately calculate the optimal fingerprint data.

[0083] S2. Searching for source fingerprint data of feedback points within the preset range of non-feedback points.

[0084] Specifically, the step S2 includes:

[0085] S21. In the initial fingerprint database, search for a target reference point whose relative position relationship with the non-feedback point is within a preset range, and mark the AP signal value and the position information of the target reference point as source fingerprint data of the feedback point.

[0086] In some embodiments, the target reference point located in the vicinity of the non-feedback point can be determined through the initial fingerprint database, and the fingerprint data of the target reference point in the initial fingerprint database is marked as the source fingerprint data of the feedback point.

[0087] S3. Based on a radial basis function, construct an estimated fingerprint surface according to the non-feedback points and the source fingerprint data.

[0088] Specifically, step S3 includes:

[0089] S31, defining a cut-off region as a circular region with the non-feedback point as the center and including the feedback point, and calculating the distance between the non-feedback point and each of the feedback points in the cut-off region.

[0090] In some embodiments, after determining the radius of the cut-off area, a circular area is delineated with the non-feedback point as the center, and feedback points are randomly and evenly selected in all directions of the circular area to ensure that the feedback points are evenly scattered in the circular area. The radius of the cut-off area can be defined according to the grid distance, or a triangle is constructed based on three or more known feedback points, the circumcenter of the triangle is the center of the circle, and the distance from the center of the circle to any point is the radius R, and then a new circular area is defined based on the existing circular area and other points.

[0091] S32. Calculate the radial basis function value of each distance by using a radial basis function.

[0092] In some embodiments, step S32 is specifically in, represents the radial basis function, and ||pq|| represents the physical distance between the non-feedback point p and the feedback point q.

[0093] S33, gridding the cutoff area based on a preset interval distance, and performing interpolation calculation on each grid point according to the radial basis function value and the source fingerprint data to obtain an estimated fingerprint surface.

[0094] In some embodiments, the cutoff region is gridded at intervals of 1 m, and each grid point is interpolated based on the radial basis value and the source fingerprint data to obtain an estimated AP signal value for each grid point, thereby obtaining the estimated AP signal values ​​of all grid points in the cutoff region, i.e., the estimated fingerprint surface.

[0095] S4. Determine an estimated AP signal value of an area corresponding to the non-feedback point according to the estimated fingerprint surface.

[0096] In some embodiments, the area corresponding to the non-feedback point is a cut-off area with the non-feedback point as the center.

[0097] S5. Input the estimated AP signal value into a pre-trained deep learning model to obtain estimated location information.

[0098] It should be noted that, since the algorithm for interpolation calculation based on radial basis function values ​​only considers the signal of one AP device each time it is executed, if the signals corresponding to all AP devices are to be obtained, the calculation needs to be repeated, which increases the calculation workload. Therefore, the present invention optimizes and screens the signals of multiple AP devices through steps S501 and S502.

[0099] In an optional implementation, before step S5, the method further includes:

[0100] S501: Optimize and screen the estimated AP signal value according to the cutoff region corresponding to each of the non-feedback points.

[0101] Specifically, the step S501 includes:

[0102] S5011. Determine an overlapping area between the cut-off areas corresponding to each of the non-feedback points.

[0103] S5012: Divide the overlapping area into grids based on the preset interval distance, and determine the coordinate position of the estimated AP signal value according to the grid points.

[0104] S5013: Select the estimated AP signal value whose coordinate position is within the overlapping area.

[0105] In an optional implementation, before step S5, the method further includes:

[0106] S502: Optimize and filter the estimated AP signal value according to the source fingerprint data.

[0107] Specifically, the source fingerprint data includes a timestamp, and step S502 includes:

[0108] S5021. Determine a first priority of the source fingerprint data according to the sequence of the timestamps.

[0109] In some embodiments, the closer the acquisition timestamp of the source fingerprint data is to the current time, the higher the priority of the source fingerprint data. Specifically, the first priority of the source fingerprint data can also be determined comprehensively based on factors affecting the acquisition accuracy such as the location where the source fingerprint data is collected and the surrounding environment.

[0110] S5022. Determine, according to the first priority, a second priority of the estimated AP signal value obtained by interpolation based on the source fingerprint data.

[0111] In some embodiments, when the estimated AP signal value is obtained based on the interpolation of the source fingerprint data, the source fingerprint data and the corresponding estimated AP signal value are marked to facilitate determining the second priority of the estimated AP signal value in step S5022.

[0112] S5023. Select the estimated AP signal value with the second priority as the highest level.

[0113] In some embodiments, after obtaining the estimated AP signal value, a signal completion enhancement mechanism can be established through an adaptive filtering model to optimize the estimated AP signal value. The adaptive filtering model monitors the characteristics and environmental conditions of the estimated AP signal value, and automatically adjusts the parameters in the filtering model based on this information to minimize the mean square error between the expected signal value and the estimated AP signal value. Specifically, a filter training sample X(n) is obtained, and the parameters of the filtering model are initialized. For a set of filter training samples X(n) and the corresponding expected output d(n), the error signal e(n) = expected output d(n) - actual output y(n) can be obtained, then the parameter W(n+1) of the filtering model = original parameter W(n) + learning rate η × error signal e(n). In this process, it is determined whether the error signal e(n) meets the convergence condition. If it meets the convergence condition, it ends. If not, it continues to iterate and update the parameters to obtain a trained adaptive filtering model. The estimated AP signal value is denoised, enhanced or repaired through the adaptive filtering model.

[0114] S6. Generate standard fingerprint data according to the estimated AP signal value and the estimated location information.

[0115] S7. Update the initial fingerprint database according to the standard fingerprint data to obtain a standard fingerprint database.

[0116] In some embodiments, after the initial fingerprint database is updated according to the standard fingerprint data, the administrator maintains and collects the current updated printer information, such as adding or deleting printer locations. Specifically, a QR code image containing printer information is attached to the printer device, and the administrator can update the printer information by scanning the code.

[0117] S8. In response to the user's printing instruction, locate the target printer according to the standard fingerprint database and the printing instruction to complete the printing operation through the target printer.

[0118] Specifically, the step S8 includes:

[0119] S81. Acquire a real-time AP signal value of a mobile terminal corresponding to the printing instruction.

[0120] S82: Determine the scene type of the real-time AP signal value based on a scene analysis model, and obtain first candidate fingerprint data from the standard fingerprint database according to the scene type.

[0121] In some embodiments, the standard fingerprint data may also include a picture of the floor where the printer is located and a picture of the printer device to avoid the situation where multiple printers on the same floor are too close to be distinguished by AP positioning, so that users can quickly determine the target printer they need.

[0122] In some embodiments, the location coordinate format of the printer in the standard fingerprint database is specifically company name-campus name-building number-floor-scene type (area) to facilitate identification and distinction of different printers.

[0123] S83: Determine, from the first candidate fingerprint data, second candidate fingerprint data closest to the real-time AP signal value by using a neighboring information model.

[0124] S84: Generate a candidate printer list according to the second candidate fingerprint data, and return it to the mobile terminal according to the print instruction.

[0125] S85. In response to the selection instruction of the mobile terminal, determine a target printer in the candidate printer list.

[0126] In some embodiments, in the target environment, in order to avoid the problem of AP cross-floor identification leading to incorrect positioning of the target printer, the AP device can actively monitor weak network devices and actively disconnect, so that the mobile terminal can reconnect to the AP device with the strongest signal to ensure that the AP device connected to the mobile terminal is the AP device on the current floor.

[0127] In some embodiments, the process of realizing cloud printing by data interaction in each module of the cloud printing system of the present invention is specifically as follows: the user sends a request to the AP device in the target environment through the mobile terminal to obtain the AP signal value of the mobile terminal at the current location. According to the obtained AP signal value, the standard fingerprint database is matched to return the location information of the printer near the current location. The open platform unified entrance of the cloud printing system forwards the request of the mobile terminal to the cloud printing open interface. The cloud printing open interface obtains the printer that meets the query conditions according to the right matching and fuzzy query rules. For example, the printer of the A scene type (area) in the A building, A floor, A park of the A company can be queried by entering the A scene type (area). The cloud printing server returns the queried printer list to the cloud printing open interface. The cloud printing open interface forwards the printer list to the open unified entrance of the cloud printing system and displays it on the mobile terminal. The user selects the target printer according to the returned printer list for printing operation. The development unified entrance of the cloud printing system responds to the user's printing instruction and transmits the printer information, printing-related data and printing interface selected by the user to the cloud printing server through the cloud printing open interface. The cloud printing server then sends the print file to the cloud printer through the controller for printing. Among them, the controller can uniformly and standardize the registration and transmission of printing data and parameters, is compatible with different printer types, avoids the subsequent increase in maintenance costs of various printer manufacturers, and better manages cloud printers.

[0128] Please refer to Figure 2 , Embodiment 2 of the present invention is:

[0129] A cloud printing terminal 100 based on AP positioning includes a memory 101, a processor 102, and a computer program stored in the memory 101 and running on the processor 102. When the processor 102 executes the computer program, each step in a cloud printing method based on AP positioning in Example 1 is implemented.

[0130] In summary, the present invention provides a cloud printing method and terminal based on AP positioning. When the printer position changes, the reference point that needs to be updated in the current environment is confirmed as a non-feedback point, and a cutoff area is constructed with the non-feedback point as the center of the circle. In the cutoff area, the distance between the non-feedback point and the feedback point and the source fingerprint data are interpolated by the radial basis function to obtain an estimated fingerprint surface. All signal values ​​in the vicinity of the non-feedback point are updated by simulating the estimated fingerprint surface. Finally, the estimated position information is inferred from the estimated AP signal value through the pre-trained deep learning model, so as to generate standard fingerprint data according to the estimated AP signal value and the estimated position information. In this way, the fingerprint data of the non-feedback point can be adaptively updated based on the fingerprint data in the old fingerprint database, without the need to re-collect large quantities of data, reducing time and labor cost overhead. At the same time, the initial fingerprint database is updated based on the standard fingerprint data of the non-feedback point to avoid the increase in error caused by updating only part of the feedback points, thereby ensuring the positioning accuracy of the subsequent target printer.

[0131] The above descriptions are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent transformations made using the contents of the present invention's specification and drawings, or directly or indirectly applied in related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A cloud printing method based on AP positioning, characterized in that: include: In response to the printer's position update instruction, obtaining the non-feedback point after the printer position is updated; Searching for source fingerprint data of a feedback point located within a preset range of the non-feedback point; Based on the radial basis function, constructing an estimated fingerprint surface according to the non-feedback points and the source fingerprint data; Determining an estimated AP signal value of an area corresponding to the non-feedback point according to the estimated fingerprint surface; Inputting the estimated AP signal value into a pre-trained deep learning model to obtain estimated location information; Generate standard fingerprint data according to the estimated AP signal value and the estimated location information; Update the initial fingerprint database according to the standard fingerprint data to obtain a standard fingerprint database; In response to a print instruction from a user, a target printer is located according to the standard fingerprint database and the print instruction, so as to complete a print operation through the target printer.

2. The method according to claim 1, characterized in that Before responding to the printer's location update instruction, the method further includes: Determine the reference point for the initial position of the printer in the target environment; Acquire the AP signal value and location information collected by the reference point; An initial fingerprint database is constructed according to the reference point, the AP signal value and the location information.

3. The method according to claim 2, characterized in that The step of searching for source fingerprint data of feedback points located within the preset range of non-feedback points includes: In the initial fingerprint database, a target reference point whose relative position relationship with the non-feedback point meets a preset range is searched, and the AP signal value and the position information of the target reference point are marked as source fingerprint data of the feedback point.

4. The method according to claim 1, characterized in that: The step of constructing the estimated fingerprint surface based on the radial basis function according to the non-feedback points and the source fingerprint data comprises: Define a cut-off area as a circular area with the non-feedback point as the center and including the feedback point, and calculate the distance between the non-feedback point and each of the feedback points in the cut-off area; Calculate the radial basis function value of each distance by using the radial basis function; The cut-off area is grid-divided based on a preset interval distance, and each grid point is interpolated according to the radial basis function value and the source fingerprint data to obtain an estimated fingerprint surface.

5. The method according to claim 4, characterized in that Before inputting the estimated AP signal value into the pre-trained deep learning model to generate standard fingerprint data, the method further includes: The estimated AP signal value is optimized and screened according to the cutoff region corresponding to each of the non-feedback points.

6. The method according to claim 5, characterized in that The optimizing and screening the estimated AP signal value according to the cut-off area corresponding to each of the non-feedback points includes: Determine an overlapping area between the cut-off areas corresponding to each of the non-feedback points; Gridding the overlapping area based on the preset interval distance, and determining the coordinate position of the estimated AP signal value according to the grid points; The estimated AP signal value whose coordinate position is within the overlapping area is selected.

7. The method according to claim 4, characterized in that Before inputting the estimated AP signal value into the pre-trained deep learning model to generate standard fingerprint data, the method further includes: The estimated AP signal value is optimized and screened according to the source fingerprint data.

8. The method according to claim 7, characterized in that The source fingerprint data includes a timestamp; The optimizing and screening the estimated AP signal value according to the source fingerprint data includes: Determining a first priority of the source fingerprint data according to a sequence of the timestamps; Determining, according to the first priority, a second priority of the estimated AP signal value obtained by interpolation based on the source fingerprint data; The estimated AP signal value with the second priority as the highest level is selected.

9. The method according to claim 1, characterized in that: The locating the target printer according to the standard fingerprint database and the printing instruction comprises: Acquire a real-time AP signal value of a mobile terminal corresponding to the printing instruction; Determine the scene type of the real-time AP signal value based on the scene analysis model, and obtain first candidate fingerprint data from the standard fingerprint database according to the scene type; Determine, from the first candidate fingerprint data, second candidate fingerprint data closest to the real-time AP signal value by using a neighboring information model; generating a candidate printer list according to the second candidate fingerprint data, and returning the list to the mobile terminal according to the printing instruction; In response to a selection instruction of the mobile terminal, a target printer is determined in the candidate printer list.

10. A cloud printing terminal based on AP positioning, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the processor implements each step in the cloud printing method based on AP positioning as described in any one of claims 1-9.

Images