Data processing system for determining abnormal power utilization state of target area

Abstract

The invention relates to a data processing system for determining an abnormal power utilization state of a target area, which comprises a processor and a memory in which a computer program is stored, and when the computer program is executed by the processor, the following steps are realized: obtaining a candidate equipment ID list set, obtaining a target characteristic value list set, obtaining a candidate characteristic vector list, and obtaining a target characteristic value list set; obtaining a key feature value list, obtaining a candidate priority list, obtaining a target prediction model in the training of the target function, inputting a target feature vector corresponding to the target region into the target prediction model to obtain a target priority, and determining the abnormal power consumption state based on the target priority. According to the embodiment, the power consumption characteristics of the region and the behavior characteristics corresponding to the equipment are combined to obtain the target prediction model, and the behavior characteristic data corresponding to the equipment are subjected to characteristic screening based on the historical data, so that the dimension of the data characteristics and the difference of the data are reduced, and the accuracy of determining the abnormal power consumption state of the target region is relatively high.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and particularly to a data processing system for determining an abnormal power consumption state of a target area. Background Art

[0002] In the social and economic development, electric energy plays a crucial role. Various researches and investigations directly link electric energy consumption with the national economy, technology, and social development. On the one hand, the demand for electric energy is increasing exponentially, and the available resources are being consumed at an alarming rate; on the other hand, electric energy is still in short supply, and energy conservation is a basic requirement. Therefore, power management should be strengthened and the use of electric energy should be optimized to reduce production costs and environmental hazards. Power consumption prediction and analysis are important means to achieve this goal.

[0003] In the prior art, the method for determining the abnormal power consumption state of a target area is as follows: capturing in real time the power consumption characteristic data corresponding to the target area (such as the change in current, the change in cable temperature), and inputting these data into an abnormal power consumption recognition model for recognition processing to determine the abnormal power consumption state corresponding to the target area.

[0004] In summary, the problems existing in the method for determining the abnormal power consumption state of a target area are as follows: being limited to power consumption characteristics, not combining power consumption characteristics with the behavior characteristics corresponding to the equipment, and not performing feature screening on the data, resulting in a large data feature dimension and large data differences, and thus the accuracy of determining the abnormal power consumption state of the target area is relatively low. Summary of the Invention

[0005] In view of the above technical problems, the technical solution adopted by the present invention is as follows: S100, obtaining a candidate device ID list set A = {A 1 , ……, A i , ……, A n} corresponding to the sample area list, where A i = {A i1 , ……, A ij , ……, A im(i)}, A ij is the j-th candidate device ID in the candidate device ID list corresponding to the i-th sample area, j = 1 …… m(i), m(i) is the number of candidate device IDs in the candidate device ID list corresponding to the i-th sample area, and i = 1 …… n, n is the number of sample areas.

[0006] S200, obtaining a target feature value list set G = {G 1 , ……, G i , ……, G n} corresponding to the sample area list, where G i = {G i1 , ……, G ij , ……, Gim(i)}, G ij = {G 1 ij , ……, G e ij , ……, G f ij}, G e ij is the e-th target eigenvalue in the target eigenvalue list corresponding to A, where e = 1... f, and f is the number of target eigenvalues in the target eigenvalue list. ij

[0007] S300. Obtain the candidate feature vector list J = {J 1 , ……, J i , ……, J n} corresponding to the sample region list according to G, where J i is the candidate feature vector corresponding to the i-th sample region. Among them, J i = (J i1 , ……, J ie , ……, J if ), and J ie meets the following conditions:

[0008] J ie = ∑ m(i) j＝1 G e ij .

[0009] S400. Obtain the key eigenvalue list H = {H 1 , ……, H i , ……, H n} corresponding to the sample region list, where H i is the key eigenvalue corresponding to the i-th sample region. Among them, the key eigenvalue is the power consumption of the sample region within a specified time period.

[0010] S500. Obtain the candidate priority list P = {P 1 , ……, P i , ……, P n} corresponding to the sample region list according to H, where P i is the candidate feature priority corresponding to the i-th sample region. Among them, P i meets the following conditions:

[0011] P i = 1 - (H i / T 0 ) / max{H i / T 0 , β i} + 1, T0 is the time span corresponding to the specified time period, and β i is the average normal power consumption during the working time period of the i-th sample area.

[0012] S600. Input J and P into the objective function for training, and obtain the parameters corresponding to the objective function to obtain the target prediction model.

[0013] S700. Input the target feature vector corresponding to the target area into the target prediction model, and obtain the target priority η corresponding to the target area.

[0014] S800. When η < η 0 it is determined that the target area is in an abnormal power consumption state during the pending time period, where η 0 is the preset priority threshold.

[0015] Compared with the prior art, the present invention has obvious beneficial effects. By means of the above technical solutions, a data processing system for determining the abnormal power consumption state of a target area provided by the present invention can achieve considerable technical progress and practicality, and has wide utilization value in the industry. It has at least the following beneficial effects:

[0016] A data processing system for determining the abnormal power consumption state of a target area includes: a processor and a memory storing a computer program. When the computer program is executed by the processor, the following steps are implemented: obtaining a list set of candidate device IDs corresponding to the sample area list, obtaining a list set of target feature values corresponding to the sample area list, obtaining a list of candidate feature vectors corresponding to the sample area list according to the list set of target feature values, obtaining a list of key feature values corresponding to the sample area list, obtaining a list of candidate priorities corresponding to the sample area list according to the list of key feature values, inputting the list of candidate feature vectors and the list of candidate priorities into the objective function for training, obtaining the parameters corresponding to the objective function to obtain the target prediction model, inputting the target feature vector corresponding to the target area into the target prediction model, obtaining the target priority corresponding to the target area, and when the target priority is less than the preset priority threshold, determining that the target area is in an abnormal power consumption state during the pending time period. It is not limited to the power consumption characteristics of the area, but considers the behavior characteristics of the devices corresponding to the area, combines the power consumption characteristics of the area and the behavior characteristics of the devices corresponding to the area to obtain the target prediction model, and performs feature screening on the behavior characteristic data corresponding to the devices based on historical data, reducing the dimension of the data characteristics and the difference of the data, so that the accuracy of determining the abnormal power consumption state of the target area is relatively high.

[0017] The above description is only an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention, it can be implemented according to the content of the specification. And in order to make the above and other purposes, features and advantages of the present invention more obvious and understandable, the following preferred embodiments are specifically given and described in detail in conjunction with the accompanying drawings. Description of the Drawings

[0018] Figure 1 It is a flowchart of a computer program executed by a data processing system for determining the abnormal power consumption state of a target area provided in an embodiment of the present invention. Detailed Embodiment

[0019] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative efforts belong to the scope of protection of the present invention.

[0020] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above drawings are used to distinguish similar objects, and do not necessarily need to describe a specific order or sequence. It should be understood that such data used can be interchanged under appropriate circumstances so that the embodiments of the present invention described here can be implemented in an order different from those illustrated or described here. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or server including a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0021] Embodiment

[0022] This embodiment provides a data processing system for determining the abnormal power consumption state of a target area. The system includes: a processor and a memory storing a computer program. When the computer program is executed by the processor, the following steps are implemented, as Figure 1 shown:

[0023] S100, obtain a list set A of candidate device IDs corresponding to the sample area list A = {A 1 , ……, A i , ……, A n}, A i = {A i1 , ……, A ij , ……, A im(i)}, A ijis the j-th candidate device ID in the list of candidate device IDs corresponding to the i-th sample area, where j = 1... m(i), m(i) is the number of candidate device IDs in the list of candidate device IDs corresponding to the i-th sample area, and i = 1... n, where n is the number of sample areas.

[0024] Specifically, the list of sample areas includes several sample areas, where the sample area is an area that aggregates several target objects and provides an office space for the target objects, such as: company office locations, certain departments of a company, and other sample areas.

[0025] Furthermore, the target object is a user who uses the candidate device.

[0026] The candidate device ID is a unique identifier representing the candidate device, where the candidate device is a user connected to the wifi in the sample area.

[0027] S200, obtain the set of target feature value lists G = {G 1 , ……, G i , ……, G n} corresponding to the list of sample areas, where G i = {G i1 , ……, G ij , ……, G im(i)}, G ij = {G 1 ij , ……, G e ij , ……, G f ij}, and G e ij is the e-th target feature value in the target feature value list corresponding to A ij , where e = 1... f and f is the number of target feature values in the target feature value list.

[0028] Specifically, the target feature value is the value of the target feature corresponding to the candidate device within a preset time period.

[0029] Specifically, in S200, the target feature is obtained through the following steps:

[0030] S201, obtain the candidate feature list D = {D 1 , ……, D r , ……, D s}, where D r is the r-th candidate feature, and r = 1... s, where s is the number of candidate features.

[0031] Specifically, the candidate feature is the behavior feature of the target object corresponding to the candidate device in the sample area, such as: the number of times the APP is opened, the duration of using the APP, the working duration, and other behavior features.

[0032] S202. Obtain the first candidate feature value list set F = {F 1 , ……, F r , ……, F s} corresponding to D, where F r = {F r1 , ……, F ri , ……, F rn} is the i-th first candidate feature value in the first candidate feature value list corresponding to D, where F ri satisfies the following conditions: r ri ri F

[0033] m(i) j＝1 = (1 / m(i)) × ∑ m(i) j＝1 C r ij / m(i), where C r ij is the value of the candidate feature corresponding to the candidate device corresponding to A within the preset time period. ij Specifically, the end time of the preset time period is earlier than the current time. Those skilled in the art know that the selection of the span of the preset time period can be made according to actual needs, all of which fall within the protection scope of the present invention and will not be elaborated here.

[0034]

[0035] S203. Obtain the second candidate feature value list E = {E 1 , ……, E i , ……, E n} corresponding to the sample area list, where E i is the second candidate feature value corresponding to the i-th sample area, and the second candidate feature value is the power consumed by the sample area within the preset time period.

[0036] S204. According to F and E, obtain the specified priority list Q = {Q 1 , ……, Q r , ……, Q s} corresponding to D, where Q r is the specified priority corresponding to D r , and Q r satisfies the following conditions:

[0037]

[0038] S205. Obtain f target features according to Q, where when |Q r | < Q 0 , determine D r as the target feature, Q 0 is the preset priority threshold, |Q r | is the absolute value of Q r .

[0039] Specifically, the value range of Q 0 is 0.3 to 0.4. Those skilled in the art know that the selection that can be made according to actual needs all fall within the protection scope of the present invention and will not be elaborated here.

[0040] As described above, obtaining the specified priority list corresponding to the candidate feature list, screening the candidate features based on the specified priority, obtaining the target features, and performing feature screening on the behavior feature data corresponding to the device based on historical data reduce the dimension of the data features and the difference of the data, making the accuracy of determining the abnormal power consumption state of the target area relatively high.

[0041] S300. Obtain the candidate feature vector list J = {J 1 , ……, J i , ……, J n} corresponding to the sample area list, where J i is the candidate feature vector corresponding to the i-th sample area, and among them, J i = (J i1 , ……, J ie , ……, J if ), and J ie meets the following conditions:

[0042] J ie = ∑ m(i) j＝1 G e ij .

[0043] S400. Obtain the key feature value list H = {H 1 , ……, H i , ……, H n} corresponding to the sample area list, where H i is the key feature value corresponding to the i-th sample area, and the key feature value is the power consumed by the sample area within the specified time period.

[0044] Specifically, the start time corresponding to the specified time period is later than the end time of the preset time period. Those skilled in the art know that the selection of the time span corresponding to the specified time period that can be made according to actual needs all fall within the protection scope of the present invention and will not be elaborated here.

[0045] S500. According to H, obtain the candidate priority list P = {P 1 , ……, P i , ……, P n} corresponding to the sample area list, where P i is the candidate feature priority corresponding to the i-th sample area. Among them, P i meets the following conditions:

[0046] P i = 1 - (H i / T 0 ) / max{H i / T 0 , β i}+1, where T 0 is the time span corresponding to the specified time period, and β i is the average normal power consumption during the working time period of the i-th sample area.

[0047] S600. Input J and P into the training of the objective function to obtain the parameters corresponding to the objective function to obtain the target prediction model.

[0048] Preferably, the objective function is the Sigmoid function.

[0049] S700. Input the target feature vector corresponding to the target area into the target prediction model to obtain the target priority η corresponding to the target area.

[0050] Specifically, the target area is the area to be confirmed whether it is in an abnormal power consumption state.

[0051] Furthermore, the nature of the target area is the same as that of the sample area.

[0052] Specifically, the target feature vector is a feature vector obtained based on a number of undetermined feature values corresponding to a number of devices in the target area, where the undetermined feature value is the value under the target feature corresponding to the device in the target area during the undetermined time period.

[0053] Furthermore, the acquisition method of the target feature vector is the same as that of the candidate feature vector.

[0054] Furthermore, the end time of the undetermined time period is earlier than the current time. It should be noted that those skilled in the art know that the time span corresponding to the undetermined time period can be selected according to actual needs, and all fall within the protection scope of the present invention, so it will not be elaborated here.

[0055] Specifically, the target priority η is the value obtained by inputting the target feature vector corresponding to the target area into the target prediction model.

[0056] S800, when η < η 0 , it is determined that the target area is in an abnormal power consumption state during the to-be-determined time period, where η 0 is a preset priority threshold.

[0057] Specifically, the smaller η is, the more abnormal the power consumption of the target area is during the to-be-determined time period. Those skilled in the art know that any selection made according to actual needs falls within the protection scope of the present invention and will not be elaborated here.

[0058] Based on the above, obtaining the candidate feature vector list corresponding to the sample area list and the key feature value list corresponding to the sample area list, obtaining the target prediction model based on the key feature vector list and the key feature value list, and determining the abnormal power consumption state corresponding to the target area based on the target prediction model. It is not limited to the power consumption characteristics of the area, but considers the behavior characteristics of the devices in the area, and combines the power consumption characteristics of the area and the behavior characteristics of the corresponding devices to obtain the target prediction model, so that the accuracy of determining the abnormal power consumption state of the target area is relatively high.

[0059] A data processing system for determining the abnormal power consumption state of a target area provided in this embodiment, the system includes: a processor and a memory storing a computer program. When the computer program is executed by the processor, the following steps are implemented: obtaining the candidate device ID list set corresponding to the sample area list, obtaining the target feature value list set corresponding to the sample area list, obtaining the candidate feature vector list corresponding to the sample area list according to the target feature value list set, obtaining the key feature value list corresponding to the sample area list, obtaining the candidate priority list corresponding to the sample area list according to the key feature value list, inputting the candidate feature vector list and the candidate priority list into the training of the target function, obtaining the parameters corresponding to the target function to obtain the target prediction model, inputting the target feature vector corresponding to the target area into the target prediction model, obtaining the target priority corresponding to the target area, and when the target priority is less than the preset priority threshold, determining that the target area is in an abnormal power consumption state during the to-be-determined time period. This embodiment is not limited to the power consumption characteristics of the area, but considers the behavior characteristics of the devices in the area, and combines the power consumption characteristics of the area and the behavior characteristics of the corresponding devices to obtain the target prediction model. Feature screening is performed on the behavior characteristic data corresponding to the devices based on historical data, reducing the dimension of the data features and the difference of the data, so that the accuracy of determining the abnormal power consumption state of the target area is relatively high.

[0060] Although some specific embodiments of the present invention have been described in detail by way of examples, those skilled in the art should understand that the above examples are for illustrative purposes only and not for limiting the scope of the present invention. Those skilled in the art should also understand that various modifications can be made to the embodiments without departing from the scope and spirit of the present invention. The scope of the present invention is defined by the appended claims.

Claims

1. A data processing system for determining abnormal power consumption status in a target area, characterized in that: The system comprises: a processor and a memory storing a computer program, and when the computer program is executed by the processor, the following steps are implemented: S100, obtaining a candidate device ID list set A corresponding to the sample area list = {A1, ..., A i , ..., A n },A i ={A i1 , ..., A ij , ..., A im(i) }, A ij is the jth candidate device ID in the candidate device ID list corresponding to the i-th sample area, j=1...m(i), m(i) is the number of candidate device IDs in the candidate device ID list corresponding to the i-th sample area, i=1...n, n is the number of sample areas; S200, obtaining a target feature value list set G corresponding to the sample area list = {G1, ..., G i , ..., G n },G i = {G i1 , ..., G ij , ..., G im(i) }, G ij = {G 1 ij , ..., G e ij , ..., G f ij }, G e ij A ij The e-th target eigenvalue in the corresponding target eigenvalue list, e=1…f, f is the number of target eigenvalues ​​in the target eigenvalue list; S300, according to G, obtain a candidate feature vector list J corresponding to the sample area list = {J1, ..., J i , ..., J n }, J i is the candidate feature vector corresponding to the i-th sample area, where J i =(J i1 , ..., J ie , ..., J if ), J ie Meet the following conditions: J ie =∑ m(i) j＝1 G e ij ; S400, obtaining a key feature value list H corresponding to the sample area list = {H1, ..., H i , ..., H n }, H i is the key characteristic value corresponding to the i-th sample area, wherein the key characteristic value is the power consumed by the sample area in a specified time period; S500, according to H, obtain a candidate priority list P corresponding to the sample area list = {P1, ..., P i , ..., P n }, P i is the candidate feature priority corresponding to the i-th sample area, where P i Meet the following conditions: P i =1-(H i / T 0 ) / max{H i / T 0 , β i }+1,T 0 is the time span corresponding to the specified time period, β i is the normal average power consumption of the i-th sample area during the working period; S600, inputting J and P into the objective function for training, obtaining parameters corresponding to the objective function to obtain a target prediction model; S700, inputting a target feature vector corresponding to the target area into a target prediction model to obtain a target priority η corresponding to the target area; S800, when η<η 0 When , it is determined that the target area is in an abnormal power consumption state during the to-be-determined time period, where η 0 is the preset priority threshold.

2. The data processing system for determining abnormal power consumption status in a target area according to claim 1, characterized in that: The sample area list includes a plurality of sample areas, wherein the sample areas are areas where a plurality of target objects are gathered and which provide office spaces for the target objects.

3. The data processing system for determining abnormal power consumption status in a target area according to claim 1, characterized in that: The target feature value is a value of the target feature corresponding to the candidate device within a preset time period.

4. The data processing system for determining abnormal power consumption status in a target area according to claim 3, characterized in that: In S200, the target features are obtained through the following steps: S201, obtain a candidate feature list D = {D1, ..., D r , ..., D s }, D r is the rth candidate feature, r=1…s, s is the number of candidate features; S202, obtain the first candidate feature value list set F corresponding to D = {F1, ..., F r , ..., F s }, F r ={F r1 , ..., F ri , ..., F rn }, F ri D r The i-th first candidate eigenvalue in the corresponding first candidate eigenvalue list, where F ri Meet the following conditions: F ri =(1 / m(i))×∑ m(i) j＝1 C r ij / m(i), C r ij A ij The value of the candidate feature of the corresponding candidate device within the preset time period; S203, obtaining a second candidate feature value list E corresponding to the sample area list = {E1, ..., E i ,……,E n }, E i is a second candidate characteristic value corresponding to the i-th sample area, wherein the second candidate characteristic value is the power consumed by the sample area within a preset time period; S204, according to F and E, obtain the specified priority list Q corresponding to D = {Q1, ..., Q r , ..., Q s }, Q r D r The corresponding specified priority, where Q r Meet the following conditions: S205, according to Q, obtain f target features, where |Q r |<Q 0 When D r is the target feature, Q 0 is the preset priority threshold, |Q r | for Q r The absolute value of .

5. The data processing system for determining abnormal power consumption status in a target area according to claim 4, characterized in that: The candidate features are behavior features of the target object corresponding to the candidate devices in the sample area.

6. The data processing system for determining abnormal power consumption status in a target area according to claim 4, characterized in that: Q 0 The value range is 0.3~0.

4.

7. The data processing system for determining abnormal power consumption status in a target area according to claim 1, characterized in that: The target feature vector is a feature vector obtained based on a number of pending feature values ​​corresponding to a number of devices in the target area, wherein the pending feature value is a value of the target feature corresponding to the device in the target area within a pending time period.

8. The data processing system for determining abnormal power consumption status in a target area according to claim 1, characterized in that: The target priority η is a value obtained by inputting the target feature vector corresponding to the target area into the target prediction model.

Images