A dynamic data collection and processing method and system based on cloud computing

Abstract

The present invention relates to the field of dynamic data processing, and specifically to a dynamic data acquisition and processing method and system based on cloud computing, comprising: a dynamic capture module, a signal modulation module, a cloud preloading module, a pulse update module and a precision feedback module, wherein the dynamic capture module is used to capture the call of dynamic variables, the signal modulation module is used to generate a multidimensional feature vector of dynamic data, the cloud preloading module is used to map the initial dynamic data in a cloud virtual space, the pulse update module is used to update the feature vector and send the update result through a pulse signal, and the precision feedback module is used to adjust the unit lattice distribution generated in the cloud. The present invention can ensure the timely update and feedback of dynamic data, improve the timeliness and reliability of the dynamic data processing process, reduce channel resource overhead, enable the sampling system to adapt to the dynamic data type, and improve the real-time and response speed of dynamic signal sampling and processing.

Description

Technical Field

[0001] The present invention relates to the field of dynamic data processing, and in particular to a dynamic data acquisition and processing method and system based on cloud computing. Background Art

[0002] Dynamic data refers to data that is updated in real time with user input or other changes when the program is running. It can reflect changes in system status or user operations in real time, achieving real-time monitoring, real-time data analysis, etc. The current dynamic data acquisition technology is composed of sampling software in the source device, which collects device data and uploads it to the data processing device.

[0003] In most cases, the source device and the processing device are far apart and need to be connected through the Internet, radio, etc. When the amount of sampled data is large, the transmission is slow, resulting in high latency in data acquisition and unable to guarantee real-time updates of dynamic data. Congested channels may also cause large communication errors and affect data quality.

[0004] In addition, dynamic data generally involves more variables, such as feedback data from different sensors, different items in the test log, multiple influencing parameters of the equipment's working status, etc. The update cycles of these variables are different. During transmission, source tags need to be added to determine the variable type, which not only increases the amount of data transmission, but also makes it more likely that problems will occur during data processing. Summary of the invention

[0005] The purpose of the present invention is to provide a cloud computing-based dynamic data collection and processing method and system to solve the problems raised in the above background technology.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions: a dynamic data acquisition and processing system based on cloud computing, comprising: a dynamic capture module, a signal modulation module, a cloud preloading module, a pulse update module and an accuracy feedback module;

[0007] The dynamic capture module is composed of a call monitoring chip and an unpacking program set in the source device, and is used to capture the call results of the source device on each dynamic variable, record the call time, call number and size of the dynamic variable when calling, and form a call link;

[0008] The signal modulation module is used to construct a multidimensional coordinate space according to the address of the dynamic variable, and arrange all call links pointing to the same dynamic variable in chronological order, generate a feature vector with the difference between the sizes of the dynamic variables in adjacent links as the horizontal coordinate and the call interval as the vertical coordinate, and synthesize the feature vectors corresponding to each dynamic variable in the multidimensional coordinate space to obtain a multidimensional feature vector of the dynamic data;

[0009] The cloud preloading module is used to send initial data from the source device to the cloud before data tracking begins, reconstruct the multi-dimensional coordinate space in the cloud, and the cloud is centered on the initial data value, generates a unit lattice on the corresponding space plane according to the normal distribution dynamic parameters, and maps the initial multi-dimensional feature vector in the virtual coordinate space to form a virtual feature vector;

[0010] The pulse update module is used to fit the updated multidimensional feature vector in the source device when the data of the source device is updated, and the dot product of the multidimensional feature vector before and after the data update is used as the pulse signal height, and the modulus length of the difference vector is used as the pulse width, and the modulus length of the difference vector is used as the pulse width, and the pulse signal is modulated and sent to the cloud. The cloud recognizes the pulse signal, and after processing the virtual feature vector, the mapping vector of each dynamic variable after the update is obtained, and the unit point closest to the end point pointed to by the mapping vector is output as the dynamic variable sampling result;

[0011] The precision feedback module is used to calculate the adjacent information entropy of adjacent pulse signals, predict the data change rate of subsequent pulses based on the adjacent information entropy of previous pulses, multiply the data change rate by the variance of the lattice as the new variance, and use the updated dynamic variable value as the new mean to regenerate the unit lattice in the cloud.

[0012] Further, the dynamic capture module includes: a chip calling unit and a variable analysis unit;

[0013] The chip calling unit is used to detect the calling of each dynamic parameter by the source device and record the size of the dynamic parameter at each call;

[0014] The variable analysis unit is used to generate a call link for each dynamic parameter in the database, wherein the data recorded in the call link includes: the call time, the number of calls and the size of the dynamic variable when calling.

[0015] Further, the signal modulation module includes: a virtual space unit, a vector mapping unit and a pulse generation unit;

[0016] The virtual space unit is used to divide the virtual operation space in the device memory, and construct a multi-dimensional coordinate space with each dynamic parameter as a dimension of the space;

[0017] The vector mapping unit is used to generate a feature vector by using the size of the dynamic variable and the sampling interval as the horizontal and vertical coordinates of the vector respectively;

[0018] The pulse generating unit is used to synthesize all the characteristic vectors in the multi-dimensional coordinate space to obtain the multi-dimensional characteristic vector, and to set a signal generating device on the information source device.

[0019] Furthermore, the cloud preloading module includes: a dot stacking unit and a cloud computing unit;

[0020] The dot matrix stacking unit is used to receive the current dynamic variable size of the source device, and generate a unit dot matrix on the plane with the dynamic variable as the mean value;

[0021] The cloud computing unit is used to divide cloud computing resources, use the computing resources to perform signal recognition, virtual space construction and vector decomposition operations, and output the computing results in the cloud.

[0022] Further, the pulse updating module includes: a pulse decoupling unit and a variable output unit;

[0023] The pulse decoupling unit is used to obtain the pulse signal sent by the information source identification and detect the height and width of the pulse signal;

[0024] The variable output unit is used to calculate the dynamic parameters before and after the update according to the information corresponding to the pulse signal, and output the updated dynamic parameters.

[0025] Further, the accuracy feedback module includes: a resource recording unit, an adjacency information unit and a density feedback unit;

[0026] The resource recording unit is used to record the decoupling results of the pulse signal and record them in the cloud database in chronological order;

[0027] The adjacent information unit is used to calculate the adjacent information entropy of adjacent pulse signals in the database, and multiply the change rate of the adjacent information entropy by a preset coefficient to obtain the data change rate;

[0028] The density feedback unit is used to reconstruct the unit lattice distribution function with the updated dynamic variable as the mean and the data change rate as the variance to form a new unit lattice.

[0029] A dynamic data collection and processing method based on cloud computing includes the following steps:

[0030] Step S1. A monitoring chip is set in the source device to obtain the calling address of the source device in real time, generate parameter coordinate axes according to the sizes of various dynamic parameters, and construct a multidimensional coordinate system by all parameter coordinate axes and the time axis;

[0031] Step S2. With the dynamic parameter value as the horizontal coordinate and the current time as the vertical coordinate, a feature vector is generated on the parameter axis and the time axis plane, and the feature vectors of the current dynamic parameters of each address are mapped to the multidimensional coordinate system and summed to obtain a multidimensional feature vector;

[0032] Step S3. The supervision chip uploads the multidimensional feature vector to the cloud, and the cloud constructs a coordinate space identical to the multidimensional coordinate system, obtains the end point coordinates of the multidimensional feature vector in the coordinate space, takes the vector modulus as the mean, and the preset parameter as the variance, and generates a normally distributed unit lattice around the end point coordinates;

[0033] Step S4. After the source device starts data tracking, whenever the dynamic parameters are called, the source device updates the multidimensional feature vector, takes the dot product of the multidimensional feature vectors before and after the update as the pulse height, and the modulus length of the difference vector as the pulse width, modulates the pulse sequence and sends it to the cloud. The cloud analyzes the pulse signal, and performs an inverse operation on the current feature vector of the cloud according to the vector restoration method to obtain the updated multidimensional feature vector;

[0034] Step S5. Select the unit point closest to the end point coordinates of the updated multidimensional feature vector from the unit lattice, output the mapping value of the unit point in each dimension of the space as the updated dynamic parameter value, and update the mean and variance of the unit lattice distribution, and regenerate the unit lattice with the end point coordinates as the center.

[0035] Further, step S1 includes:

[0036] Step S11. Install a supervision chip in the source device, the supervision chip is used to capture the data call pointer of the source device, detect the call address of the source device for each dynamic parameter, and read the dynamic parameter value corresponding to the address at the call address;

[0037] Step S12. Before data tracing, the supervisory chip accesses all dynamic parameter addresses in the source device to obtain the initial values ​​of the dynamic parameters, which are recorded as {W1, W2, ..., Wn}, where n represents the total number of dynamic parameter addresses and Wn represents the initial value of the dynamic parameter in the nth address;

[0038] Step S13. Construct a parameter coordinate axis with the initial value of the dynamic parameter as zero and the value increasing direction as the positive direction, construct a time axis with the current moment as zero and the time flow direction as the positive direction, and construct an n+1-dimensional coordinate system by the parameter coordinate axes of the dynamic parameters in all addresses and the time axis, recorded as a multidimensional coordinate system.

[0039] Further, step S2 includes:

[0040] Step S21. For each dynamic parameter, generate a feature vector Y on a plane formed by the parameter coordinate axis and the time axis, where Y=(c, t), where c is the dynamic parameter value and t is the current time;

[0041] Step S22. Map the feature vectors corresponding to each dynamic parameter into a multidimensional coordinate system, add all the mapping vectors to obtain a multidimensional feature vector YS, where YS=Y1+Y2+…+Yn, where Y1, Y2,…Yn represent the mapping vectors of the feature vectors corresponding to the 1st, 2nd,…nth dynamic parameters in the multidimensional coordinate system, respectively.

[0042] Further, step S3 includes:

[0043] Step S31. constructing a virtual space with the same dimension as the multi-dimensional coordinate system in the cloud, wherein the virtual space includes a visualization space and a closed space;

[0044] Step S32. The source device sends the coordinate representation of the multi-dimensional feature vector YS to the cloud. After the cloud obtains the coordinate of the vector YS, it projects the vector YS into the virtual space, performs mapping operations on each dimension of the vector YS in the virtual space, obtains each dynamic parameter value and update time, and outputs all the obtained dynamic parameter value information and update time to the sink device;

[0045] Step S33. Generate m unit points with the end point coordinates of vector YS as the center, where m is a preset value representing data accuracy. The distance and angle between each unit point and the end point coordinates satisfy the unit point lattice distribution equation:

[0046] ;

[0047] Where ai represents the distance between the i-th unit point and the end point coordinates, θi represents the angle between the i-th unit point and the end point coordinates relative to the direction of the vector YS, |YS| represents the modulus of the vector YS, i is the unit point number, and e is the base of the natural logarithm;

[0048] Step S34: The cloud marks all the obtained unit points in the virtual space to obtain a unit point matrix.

[0049] Further, step S4 includes:

[0050] Step S41. When the dynamic parameters in the source device change, the supervision chip obtains the change amount and change time of each dynamic parameter, re-determines the feature vector Y with the updated dynamic parameters, and re-executes step S22 according to the feature vectors after all dynamic parameters change within the preset period to obtain an updated multi-dimensional feature vector YT;

[0051] Step S42. The source device generates a pulse signal, the height H of the pulse signal satisfies H=YS·YT, and the width E satisfies E=|YS-YT|, and sends the pulse signal to the cloud;

[0052] Step S43. The cloud detects the pulse signal to obtain the height and width of the pulse signal, and performs an inverse operation on the current feature vector of the cloud according to the vector restoration method to obtain an updated multi-dimensional feature vector. The vector restoration method is as follows: List the first set of equations

[0053] ;

[0054] Wherein, x1, x2, ...xn represent the 1st, 2nd, ...nth coordinate values ​​of vector YT respectively, and r1, r2, ...rn represent the 1st, 2nd, ...nth coordinate values ​​of vector YS respectively;

[0055] Cloud Traversal x 1 、x 2 ,…x n The data value of x 1 、x 2 ,…x n Satisfy the first set of equations and output the first calculated value of x1, x2, ...xn in the cloud as the coordinate output of the updated vector YT.

[0056] Further, step S5 includes:

[0057] Step S51. Obtain the unit point matrix corresponding to point (x 1, x 2, …,x n ) The nearest unit point is taken as the target point, and the plane composed of the parameter coordinate axis and the time axis of each dynamic parameter is taken as the mapping plane. The target point is mapped to each mapping plane of the cloud virtual space to obtain the mapping coordinates. The horizontal coordinate value of the mapping coordinate is output as the updated value of the dynamic parameter, and the vertical coordinate value is output as the updated time of the dynamic parameter;

[0058] Step S52. Calculate the adjacent information entropy of adjacent pulse signals in the database, multiply the change rate of the adjacent information entropy by the preset coefficient to obtain the data change rate, take the module length of vector YT as the mean and the data change rate as the variance to reconstruct the unit lattice distribution equation, and update the distribution of the unit lattice.

[0059] Compared with the prior art, the beneficial effects achieved by the present invention are:

[0060] The present invention sets a call capture chip in the source device to capture the dynamic variables corresponding to each call information, generates a feature vector according to the dynamic variables, and constructs a multi-dimensional mapping space to convert the dynamic into a vector form, thereby establishing the continuity of the previous and next data, facilitating the subsequent update of the data, and improving the real-time and response speed of signal sampling.

[0061] The present invention generates a pulse signal according to the change of the eigenvector before and after the change of the eigenvector, generates a unit lattice analysis pulse signal by the cloud, calibrates the vector obtained by processing the pulse signal, and feeds back the change type and amount of the dynamic data, so that the cloud computing resources can process the data in real time through a fast connection, ensure the timely update and feedback of the dynamic data, improve the timeliness and reliability of the dynamic data processing process, and reduce the channel resource overhead.

[0062] The present invention calculates the adjacent information entropy of adjacent pulse signals and predicts the data change rate of subsequent pulses according to the adjacent information entropy of preceding pulses, thereby adjusting the generation density of the dot matrix. The change of dynamic data can be predicted in advance, and the sampling accuracy can be adjusted according to the prediction result, so that the sampling system can adapt to different data types and realize intelligent data processing according to the change mode of the data characteristics. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0064] Figure 1 It is a structural schematic diagram of a dynamic data acquisition and processing system based on cloud computing of the present invention;

[0065] Figure 2 It is a schematic diagram of the steps of a dynamic data collection and processing method based on cloud computing of the present invention. DETAILED DESCRIPTION

[0066] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0067] See also Figure 1 ,The present invention provides a technical solution: a dynamic data acquisition and processing system based on cloud computing, including: a dynamic capture module, a signal modulation module, a cloud preloading module, a pulse update module and an accuracy feedback module;

[0068] The dynamic capture module is composed of a call monitoring chip and an unpacking program set in the source device, and is used to capture the call results of the source device on each dynamic variable, record the call time, call number and size of the dynamic variable when calling, and form a call link;

[0069] The dynamic capture module includes: a chip calling unit and a variable analysis unit;

[0070] The chip calling unit is used to detect the calling of each dynamic parameter by the source device and record the size of the dynamic parameter at each call;

[0071] The variable analysis unit is used to generate a call link for each dynamic parameter in the database, wherein the data recorded in the call link includes: the call time, the number of calls and the size of the dynamic variable when calling.

[0072] The signal modulation module is used to construct a multidimensional coordinate space according to the address of the dynamic variable, and arrange all call links pointing to the same dynamic variable in chronological order, generate a feature vector with the difference between the sizes of the dynamic variables in adjacent links as the horizontal coordinate and the call interval as the vertical coordinate, and synthesize the feature vectors corresponding to each dynamic variable in the multidimensional coordinate space to obtain a multidimensional feature vector of the dynamic data;

[0073] The signal modulation module includes: a virtual space unit, a vector mapping unit and a pulse generation unit;

[0074] The virtual space unit is used to divide the virtual operation space in the device memory, and construct a multi-dimensional coordinate space with each dynamic parameter as a dimension of the space;

[0075] The vector mapping unit is used to generate a feature vector by using the size of the dynamic variable and the sampling interval as the horizontal and vertical coordinates of the vector respectively;

[0076] The pulse generating unit is used to synthesize all the characteristic vectors in the multi-dimensional coordinate space to obtain the multi-dimensional characteristic vector, and to set a signal generating device on the information source device.

[0077] The cloud preloading module is used to send initial data from the source device to the cloud before data tracking begins, reconstruct the multi-dimensional coordinate space in the cloud, and the cloud is centered on the initial data value, generates a unit lattice on the corresponding space plane according to the normal distribution dynamic parameters, and maps the initial multi-dimensional feature vector in the virtual coordinate space to form a virtual feature vector;

[0078] The cloud preloading module includes: a dot matrix stacking unit and a cloud computing unit;

[0079] The dot matrix stacking unit is used to receive the current dynamic variable size of the source device, and generate a unit dot matrix on the plane with the dynamic variable as the mean value;

[0080] The cloud computing unit is used to divide cloud computing resources, use the computing resources to perform signal recognition, virtual space construction and vector decomposition operations, and output the computing results in the cloud.

[0081] The pulse update module is used to fit the updated multidimensional feature vector in the source device when the data of the source device is updated, and the dot product of the multidimensional feature vector before and after the data update is used as the pulse signal height, and the modulus length of the difference vector is used as the pulse width, and the modulus length of the difference vector is used as the pulse width, and the pulse signal is modulated and sent to the cloud. The cloud recognizes the pulse signal, and after processing the virtual feature vector, the mapping vector of each dynamic variable after the update is obtained, and the unit point closest to the end point pointed to by the mapping vector is output as the dynamic variable sampling result;

[0082] The pulse updating module comprises: a pulse decoupling unit and a variable output unit;

[0083] The pulse decoupling unit is used to obtain the pulse signal sent by the information source identification and detect the height and width of the pulse signal;

[0084] The variable output unit is used to calculate the dynamic parameters before and after the update according to the information corresponding to the pulse signal, and output the updated dynamic parameters.

[0085] The precision feedback module is used to calculate the adjacent information entropy of adjacent pulse signals, predict the data change rate of subsequent pulses based on the adjacent information entropy of previous pulses, multiply the data change rate by the variance of the lattice as the new variance, and use the updated dynamic variable value as the new mean to regenerate the unit lattice in the cloud.

[0086] The precision feedback module includes: a resource recording unit, an adjacency information unit and a density feedback unit;

[0087] The resource recording unit is used to record the decoupling results of the pulse signal and record them in the cloud database in chronological order;

[0088] The adjacent information unit is used to calculate the adjacent information entropy of adjacent pulse signals in the database, and multiply the change rate of the adjacent information entropy by a preset coefficient to obtain the data change rate;

[0089] The density feedback unit is used to reconstruct the unit lattice distribution function with the updated dynamic variable as the mean and the data change rate as the variance to form a new unit lattice.

[0090] like Figure 2 As shown, a dynamic data collection and processing method based on cloud computing includes the following steps:

[0091] Step S1. A monitoring chip is set in the source device to obtain the calling address of the source device in real time, generate parameter coordinate axes according to the sizes of various dynamic parameters, and construct a multidimensional coordinate system by all parameter coordinate axes and the time axis;

[0092] Step S1 includes:

[0093] Step S11. Install a supervision chip in the source device, the supervision chip is used to capture the data call pointer of the source device, detect the call address of the source device for each dynamic parameter, and read the dynamic parameter value corresponding to the address at the call address;

[0094] Step S12. Before data tracing, the supervisory chip accesses all dynamic parameter addresses in the source device to obtain the initial values ​​of the dynamic parameters, which are recorded as {W1, W2, ..., Wn}, where n represents the total number of dynamic parameter addresses and Wn represents the initial value of the dynamic parameter in the nth address;

[0095] Step S13. Construct a parameter coordinate axis with the initial value of the dynamic parameter as zero and the value increasing direction as the positive direction, construct a time axis with the current moment as zero and the time flow direction as the positive direction, and construct an n+1-dimensional coordinate system by the parameter coordinate axes of the dynamic parameters in all addresses and the time axis, recorded as a multidimensional coordinate system.

[0096] Step S2. With the dynamic parameter value as the horizontal coordinate and the current time as the vertical coordinate, a feature vector is generated on the parameter axis and the time axis plane, and the feature vectors of the current dynamic parameters of each address are mapped to the multidimensional coordinate system and summed to obtain a multidimensional feature vector;

[0097] Step S2 includes:

[0098] Step S21. For each dynamic parameter, generate a feature vector Y on a plane formed by the parameter coordinate axis and the time axis, where Y=(c, t), where c is the dynamic parameter value and t is the current time;

[0099] Step S22. Map the feature vectors corresponding to each dynamic parameter into a multidimensional coordinate system, add all the mapping vectors to obtain a multidimensional feature vector YS, where YS=Y1+Y2+…+Yn, where Y1, Y2,…Yn represent the mapping vectors of the feature vectors corresponding to the 1st, 2nd,…nth dynamic parameters in the multidimensional coordinate system, respectively.

[0100] Step S3. The supervision chip uploads the multidimensional feature vector to the cloud, and the cloud constructs a coordinate space identical to the multidimensional coordinate system, obtains the end point coordinates of the multidimensional feature vector in the coordinate space, takes the vector modulus as the mean, and the preset parameter as the variance, and generates a normally distributed unit lattice around the end point coordinates;

[0101] Step S3 includes:

[0102] Step S31. constructing a virtual space with the same dimension as the multi-dimensional coordinate system in the cloud, wherein the virtual space includes a visualization space and a closed space;

[0103] Step S32. The source device sends the coordinate representation of the multi-dimensional feature vector YS to the cloud. After the cloud obtains the coordinate of the vector YS, it projects the vector YS into the virtual space, performs mapping operations on each dimension of the vector YS in the virtual space, obtains each dynamic parameter value and update time, and outputs all the obtained dynamic parameter value information and update time to the sink device;

[0104] Step S33. Generate m unit points with the end point coordinates of vector YS as the center, where m is a preset value representing data accuracy. The distance and angle between each unit point and the end point coordinates satisfy the unit point lattice distribution equation:

[0105] ;

[0106] Where ai represents the distance between the i-th unit point and the end point coordinates, θi represents the angle between the i-th unit point and the end point coordinates relative to the direction of the vector YS, |YS| represents the modulus of the vector YS, i is the unit point number, and e is the base of the natural logarithm;

[0107] Step S34: The cloud marks all the obtained unit points in the virtual space to obtain a unit point matrix.

[0108] Step S4. After the source device starts data tracking, whenever the dynamic parameters are called, the source device updates the multidimensional feature vector, takes the dot product of the multidimensional feature vectors before and after the update as the pulse height, and the modulus length of the difference vector as the pulse width, modulates the pulse sequence and sends it to the cloud. The cloud analyzes the pulse signal, and performs an inverse operation on the current feature vector of the cloud according to the vector restoration method to obtain the updated multidimensional feature vector;

[0109] Step S4 includes:

[0110] Step S41. When the dynamic parameters in the source device change, the supervision chip obtains the change amount and change time of each dynamic parameter, re-determines the feature vector Y with the updated dynamic parameters, and re-executes step S22 according to the feature vectors after all dynamic parameters change within the preset period to obtain an updated multi-dimensional feature vector YT;

[0111] Step S42. The source device generates a pulse signal, the height H of the pulse signal satisfies H=YS·YT, and the width E satisfies E=|YS-YT|, and sends the pulse signal to the cloud;

[0112] Step S43. The cloud detects the pulse signal to obtain the height and width of the pulse signal, and performs an inverse operation on the current feature vector of the cloud according to the vector restoration method to obtain an updated multi-dimensional feature vector. The vector restoration method is as follows: List the first set of equations

[0113] ;

[0114] Wherein, x1, x2, ...xn represent the 1st, 2nd, ...nth coordinate values ​​of vector YT respectively, and r1, r2, ...rn represent the 1st, 2nd, ...nth coordinate values ​​of vector YS respectively;

[0115] Cloud Traversal x 1 、x 2 ,…x n The data value of x 1 、x 2 ,…x n Satisfy the first set of equations and output the first calculated value of x1, x2, ...xn in the cloud as the coordinate output of the updated vector YT.

[0116] Step S5. Select the unit point closest to the end point coordinates of the updated multidimensional feature vector from the unit lattice, output the mapping value of the unit point in each dimension of the space as the updated dynamic parameter value, and update the mean and variance of the unit lattice distribution, and regenerate the unit lattice with the end point coordinates as the center.

[0117] Step S5 includes:

[0118] Step S51. Obtain the unit point matrix corresponding to point (x 1, x 2, …,x n ) The nearest unit point is taken as the target point, and the plane composed of the parameter coordinate axis and the time axis of each dynamic parameter is taken as the mapping plane. The target point is mapped to each mapping plane of the cloud virtual space to obtain the mapping coordinates. The horizontal coordinate value of the mapping coordinate is output as the updated value of the dynamic parameter, and the vertical coordinate value is output as the updated time of the dynamic parameter;

[0119] Step S52. Calculate the adjacent information entropy of adjacent pulse signals in the database, multiply the change rate of the adjacent information entropy by the preset coefficient to obtain the data change rate, take the module length of vector YT as the mean and the data change rate as the variance to reconstruct the unit lattice distribution equation, and update the distribution of the unit lattice.

[0120] Example: There are three dynamic parameters in a dynamic data, the initial values ​​are 1, 2 and 3 respectively, and the last update time is 11:00, 12:00 and 15:00 respectively. Then the feature vectors are (1, 11), (2, 12) and (3, 15) respectively. All feature vectors are added in four-dimensional space to obtain a multi-dimensional feature vector TS = (1.45, 1.93, 2.04, 12.7), and the vector TS is output to the cloud;

[0121] Dynamic parameters 1 and 2 change to 3 and 4 at 16:15 and 16:30 respectively, so the newly obtained multidimensional feature vector TY = (2.16, 3.03, 5.11, 15.4). The pulse signal is sent to the cloud with the dot product of TS and TY as the pulse height and the modulus of the difference vector of TS and TY as the pulse width. After the cloud obtains the pulse signal, the value of TY is calculated, and the unit point closest to TY is mapped to each plane to restore the values ​​and update time of the three dynamic parameters.

[0122] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0123] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein by equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A dynamic data collection and processing method based on cloud computing, characterized in that: The method comprises the following steps: Step S1. A monitoring chip is set in the source device to obtain the calling address of the source device in real time, generate parameter coordinate axes according to the sizes of various dynamic parameters, and construct a multidimensional coordinate system by all parameter coordinate axes and the time axis; Step S2. With the dynamic parameter value as the horizontal coordinate and the current time as the vertical coordinate, a feature vector is generated on the parameter axis and the time axis plane, and the feature vectors of the current dynamic parameters of each address are mapped to the multidimensional coordinate system and summed to obtain a multidimensional feature vector; Step S3. The supervision chip uploads the multidimensional feature vector to the cloud, and the cloud constructs a coordinate space identical to the multidimensional coordinate system, obtains the end point coordinates of the multidimensional feature vector in the coordinate space, takes the vector modulus as the mean, and the preset parameter as the variance, and generates a normally distributed unit lattice around the end point coordinates; Step S4. After the source device starts data tracking, whenever the dynamic parameters are called, the source device updates the multidimensional feature vector, takes the dot product of the multidimensional feature vectors before and after the update as the pulse height, and the modulus length of the difference vector as the pulse width, modulates the pulse sequence and sends it to the cloud. The cloud analyzes the pulse signal, and performs an inverse operation on the current feature vector of the cloud according to the vector restoration method to obtain the updated multidimensional feature vector; Step S5. Select the unit point closest to the end point coordinate of the updated multidimensional feature vector from the unit lattice, output the mapping value of the unit point in each dimension of the space as the updated dynamic parameter value, and update the mean and variance of the unit lattice distribution, and regenerate the unit lattice with the end point coordinate as the center; Step S3 includes: Step S31. constructing a virtual space with the same dimension as the multi-dimensional coordinate system in the cloud, wherein the virtual space includes a visualization space and a closed space; Step S32. The source device sends the coordinate representation of the multi-dimensional feature vector YS to the cloud. After the cloud obtains the coordinate of the vector YS, it projects the vector YS into the virtual space, performs mapping operations on each dimension of the vector YS in the virtual space, obtains each dynamic parameter value and update time, and outputs all the obtained dynamic parameter value information and update time to the sink device; Step S33. Generate m unit points with the end point coordinates of vector YS as the center, where m is a preset value representing data accuracy. The distance and angle between each unit point and the end point coordinates satisfy the unit point lattice distribution equation: ， Where ai represents the distance between the i-th unit point and the end point coordinates, θi represents the angle between the i-th unit point and the end point coordinates relative to the direction of the vector YS, |YS| represents the modulus of the vector YS, i is the unit point number, and e is the base of the natural logarithm; Step S34: The cloud marks all the obtained unit points in the virtual space to obtain a unit point matrix.

2. The method for dynamic data collection and processing based on cloud computing according to claim 1, characterized in that: Step S1 includes: Step S11. Install a supervision chip in the source device, the supervision chip is used to capture the data call pointer of the source device, detect the call address of the source device for each dynamic parameter, and read the dynamic parameter value corresponding to the address at the call address; Step S12. Before data tracing, the supervisory chip accesses all dynamic parameter addresses in the source device to obtain the initial values ​​of the dynamic parameters, which are recorded as {W1, W2, ..., Wn}, where n represents the total number of dynamic parameter addresses and Wn represents the initial value of the dynamic parameter in the nth address; Step S13. The parameter coordinate axis is constructed with the initial value of the dynamic parameter as zero and the value increasing direction as the positive direction, and the time axis is constructed with the current time as zero and the time flow direction as the positive direction. The n+1-dimensional coordinate system is constructed by the parameter coordinate axes of the dynamic parameters in all addresses and the time axis, which is recorded as a multidimensional coordinate system; Step S2 includes: Step S21. For each dynamic parameter, generate a feature vector Y on a plane formed by the parameter coordinate axis and the time axis, where Y=(c, t), where c is the dynamic parameter value and t is the current time; Step S22. Map the feature vectors corresponding to each dynamic parameter into a multidimensional coordinate system, add all the mapping vectors to obtain a multidimensional feature vector YS, where YS=Y1+Y2+…+Yn, where Y1, Y2,…Yn represent the mapping vectors of the feature vectors corresponding to the 1st, 2nd,…nth dynamic parameters in the multidimensional coordinate system, respectively.

3. The method for dynamic data collection and processing based on cloud computing according to claim 2, characterized in that: Step S4 includes: Step S41. When the dynamic parameters in the source device change, the supervision chip obtains the change amount and change time of each dynamic parameter, re-determines the feature vector Y with the updated dynamic parameters, and re-executes step S22 according to the feature vectors after all dynamic parameters change within the preset period to obtain an updated multi-dimensional feature vector YT; Step S42. The source device generates a pulse signal, the height H of the pulse signal satisfies H=YS·YT, and the width E satisfies E=|YS-YT|, and sends the pulse signal to the cloud; Step S43. The cloud detects the pulse signal to obtain the height and width of the pulse signal, and performs an inverse operation on the current feature vector of the cloud according to the vector restoration method to obtain an updated multi-dimensional feature vector. The vector restoration method is as follows: List the first set of equations ， Among them, x1, x2, …x n Represent the 1st, 2nd, ...nth coordinate values ​​of vector YT, r1, r2, ...r n Respectively represent the 1st, 2nd, ...nth coordinate values ​​of vector YS; Cloud traversal x1, x2, ...x n data values ​​so that x1, x2, …x n Satisfy the first set of equations and output the first calculated value of x1, x2, ...xn in the cloud as the coordinate output of the updated vector YT.

4. The method for dynamic data collection and processing based on cloud computing according to claim 3, characterized in that: Step S5 includes: Step S51. Obtain the unit point matrix corresponding to point (x 1, x 2, …,x n ) The nearest unit point is taken as the target point, and the plane composed of the parameter coordinate axis and the time axis of each dynamic parameter is taken as the mapping plane. The target point is mapped to each mapping plane of the cloud virtual space to obtain the mapping coordinates. The horizontal coordinate value of the mapping coordinate is output as the updated value of the dynamic parameter, and the vertical coordinate value is output as the updated time of the dynamic parameter; Step S52. Calculate the adjacent information entropy of adjacent pulse signals in the database, multiply the change rate of the adjacent information entropy by the preset coefficient to obtain the data change rate, take the module length of vector YT as the mean and the data change rate as the variance to reconstruct the unit lattice distribution equation, and update the distribution of the unit lattice.

5. A dynamic data collection and processing system based on cloud computing, characterized in that: The system includes the following modules: a dynamic capture module, a signal modulation module, a cloud preloading module, a pulse update module and an accuracy feedback module; The dynamic capture module is composed of a call monitoring chip and an unpacking program set in the source device, and is used to capture the call results of the source device on each dynamic variable, record the call time, call number and size of the dynamic variable when calling, and form a call link; The signal modulation module is used to construct a multidimensional coordinate space according to the address of the dynamic variable, and arrange all call links pointing to the same dynamic variable in chronological order, generate a feature vector with the difference between the sizes of the dynamic variables in adjacent links as the horizontal coordinate and the call interval as the vertical coordinate, and synthesize the feature vectors corresponding to each dynamic variable in the multidimensional coordinate space to obtain a multidimensional feature vector of the dynamic data; The cloud preloading module is used to send initial data from the source device to the cloud before data tracking begins, reconstruct the multi-dimensional coordinate space in the cloud, and the cloud is centered on the initial data value, generates a unit lattice on the corresponding space plane according to the normal distribution dynamic parameters, and maps the initial multi-dimensional feature vector in the virtual coordinate space to form a virtual feature vector; The pulse update module is used to fit the updated multidimensional feature vector in the source device when the data of the source device is updated, and the dot product of the multidimensional feature vector before and after the data update is used as the pulse signal height, and the modulus length of the difference vector is used as the pulse width, and the modulus length of the difference vector is used as the pulse width, and the pulse signal is modulated and sent to the cloud. The cloud recognizes the pulse signal, and after processing the virtual feature vector, the mapping vector of each dynamic variable after the update is obtained, and the unit point closest to the end point pointed to by the mapping vector is output as the dynamic variable sampling result; The precision feedback module is used to calculate the adjacent information entropy of adjacent pulse signals, predict the data change rate of subsequent pulses according to the adjacent information entropy of the preceding pulses, multiply the data change rate by the variance of the lattice as a new variance, and use the updated dynamic variable value as a new mean value to regenerate the unit lattice in the cloud; The cloud preloading module includes: a dot matrix stacking unit and a cloud computing unit; The dot matrix stacking unit is used to receive the current dynamic variable size of the source device, and generate a unit dot matrix on the plane with the dynamic variable as the mean value; The cloud computing unit is used to divide cloud computing resources, use the computing resources to perform signal recognition, virtual space construction and vector decomposition operations, and output the computing results in the cloud; The pulse updating module comprises: a pulse decoupling unit and a variable output unit; The pulse decoupling unit is used to obtain the pulse signal sent by the information source identification and detect the height and width of the pulse signal; The variable output unit is used to calculate the dynamic parameters before and after the update according to the information corresponding to the pulse signal, and output the updated dynamic parameters.

6. The cloud computing-based dynamic data collection and processing system according to claim 5, characterized in that: The dynamic capture module includes: a chip calling unit and a variable analysis unit; The chip calling unit is used to detect the calling of each dynamic parameter by the source device and record the size of the dynamic parameter at each call; The variable analysis unit is used to generate a call link for each dynamic parameter in the database, wherein the data recorded in the call link includes: the call time, the number of calls and the size of the dynamic variable when calling.

7. The cloud computing-based dynamic data collection and processing system according to claim 6, characterized in that: The signal modulation module includes: a virtual space unit, a vector mapping unit and a pulse generation unit; The virtual space unit is used to divide the virtual operation space in the device memory, and construct a multi-dimensional coordinate space with each dynamic parameter as a dimension of the space; The vector mapping unit is used to generate a feature vector by using the size of the dynamic variable and the sampling interval as the horizontal and vertical coordinates of the vector respectively; The pulse generating unit is used to synthesize all the characteristic vectors in the multi-dimensional coordinate space to obtain the multi-dimensional characteristic vector, and to set a signal generating device on the information source device.

8. The cloud computing-based dynamic data collection and processing system according to claim 7, characterized in that: The precision feedback module includes: a resource recording unit, an adjacency information unit and a density feedback unit; The resource recording unit is used to record the decoupling results of the pulse signal and record them in the cloud database in chronological order; The adjacent information unit is used to calculate the adjacent information entropy of adjacent pulse signals in the database, and multiply the change rate of the adjacent information entropy by a preset coefficient to obtain the data change rate; The density feedback unit is used to reconstruct the unit lattice distribution function with the updated dynamic variable as the mean and the data change rate as the variance to form a new unit lattice.

Images