CN111080142B - An Active Service Auxiliary Judgment Method Based on Power Failure Report - Google Patents

Abstract

The invention discloses an active service auxiliary judgment method based on power failure reporting, which comprises the following steps: a. analyzing the dimensionality influencing the user priority ranking; b. capturing historical behavior data of a user; c. establishing a user priority ranking model; d. and the user priority evaluation is realized through the historical behavior data of the user and the user priority ranking model. The invention is based on the user priority ranking algorithm of the power failure fault reporting behavior, evaluates the user priority by analyzing the historical power failure fault reporting behavior of the user, preferentially solves the problem of power failure of the user who really has a fault and is urgent, improves the user experience and improves the service quality of the power grid.

Description

A kind of active service auxiliary determination method based on power failure report

technical field

The invention belongs to the technical field of power data analysis, and in particular relates to an active service auxiliary determination method based on power failure reporting.

Background technique

With the development of Internet technology in the power industry, more and more convenient power systems are being applied online. For example, the distribution network emergency repair system can send orders for emergency repairs to users when they report a power failure. However, when a user reports a power outage, the original distribution network emergency repair system is arranged in the order of the user's access to the phone. Some false alarms and random alarms waste a lot of human resources, resulting in some users waiting too long, too many complaints, and poor user experience. And other issues.

SUMMARY OF THE INVENTION

Based on the above-mentioned deficiencies of the prior art, the present invention provides an active service assistance determination method based on power failure reporting.

The present invention is achieved through the following technical solutions.

An active service assistance determination method based on power failure reporting, comprising the following steps:

a. Analyze the dimensions that affect user prioritization;

b. Capture user historical behavior data;

c. Establish a user priority ranking model;

d. Through the user historical behavior data and the user priority sorting model, the user priority evaluation is realized.

Preferably, the dimension analysis that affects the user's prioritization in step a is as follows:

a1. The influence of the correct rate of faults reported by users on the priority of users;

a2. The impact of user complaints on user priority;

a3. The impact of user false positives on user priority;

a4. The influence of the number of calls made by the user for the same fault on the priority.

Preferably, the user historical behavior data to be captured in step b is as follows:

b1. Take the year as the unit, capture the behavior data of users reporting power outages within one year of history;

b2. Capture the fault data reported by users;

b3. Capture the data confirmed as the fault in the fault reported by the user;

b4. Capture the total number of calls reported by users for faults;

b5. Capture the number of user complaints;

b6. Capture the number of user falsely reported faults;

b7. Capture the number of calls reported by the user for this fault.

Preferably, the dimensions of the user priority ranking model in step c are as follows:

c1. The accuracy of the fault reported by the user;

c2. Frequency of incoming calls reported by users for faults;

c3. The probability of user complaints;

c4. The probability of user false positives.

Preferably, a user priority ranking model is established using the above-mentioned accuracy rate of faults reported by users, the frequency of incoming calls reported by users, the probability of user complaints, and the probability of user false alarms, as follows:

Failure Accuracy:

x=∑ _t∈1year A _t /∑ _t∈1vear B _t

Frequency of incoming calls reported by users:

y=(∑ _t∈1year T _t /∑ _t∈1year A _t ) ²

User false positive probability:

Probability of user complaints:

P(C)=(Σ _t∈1year C _t /Σ _t∈1year T _t )·E

Among them, coefficient A is the total number of reported faults, coefficient B is the number of reported faults confirmed as real faults, coefficient T is the total number of incoming calls for faults, coefficient C is the number of complaints, coefficient D is the number of falsely reported faults, and coefficient E is the number of faults reported for this time. The number of faulty calls, ∑ _t∈1year ρ _t means one year is the statistical time, where t represents time, t∈1year means the statistical time is one year; according to user habits, the priority sorting model determines two parameters, parameter 1 is F (n):

F(n)=x+y+z,

F(n)=∑ _t∈1year (A ³ +BT ² +ABD ² ) _t /∑ _t∈1year (BA ² ) _t ,

Among them, the values of T and D have a greater impact on the parameter F(n), that is, the more the number of historical false alarms, the more frequent repair reports for the same fault, and the lower the ranking of user priorities;

Parameter 2 is the user complaint probability P(C).

Preferably, according to the established user priority ranking model, two parameters F(n) and P(C) are calculated and obtained, and these two parameters are used to determine the user priority ranking, as follows:

If the parameter F(n)≤5, the user priority is rated as A;

If parameter 5<F(n)≤10, the user priority is rated as B level;

If the parameter 10<F(n)≤15, the user priority is rated as C level;

If parameter 15<F(n), the user priority is rated as D;

The parameter P(C) is used as the second criterion for user priority evaluation. When the parameter P(C)<1, the user classification is determined by the result of the parameter F(n); when the parameter P(C)≥1, directly The user is rated A.

The present invention is based on the user priority sorting algorithm based on the power failure reporting behavior. By analyzing the user's historical power failure reporting behavior for repair, the user priority is evaluated, and the real fault and the more urgent user power failure problem is solved first, and the user experience is improved. , improve the quality of grid service.

Description of drawings

Fig. 1 is the flow chart of the present invention;

FIG. 2 is a logic analysis diagram of the present invention.

Detailed ways

The present invention will be further described below in conjunction with Fig. 1 and Fig. 2:

As shown in Figure 1, an active service assistance determination method based on power failure report includes the following steps:

a. Analyze the dimensions that affect user prioritization;

b. Capture user historical behavior data;

c. Establish a user priority ranking model;

d. Through the user historical behavior data and the user priority sorting model, the user priority evaluation is realized.

Among them, the dimensions that affect user priority sorting in step a are divided into the accuracy rate of faults reported by users, user false positives, user complaints, and the frequency of incoming calls for the same fault; more false positives will reduce the priority of users; The high frequency of incoming calls for the same fault and the repeated occupation of the incoming call channel will also reduce the priority of the user, and the high accuracy of the user reporting the fault will increase the priority of the user. If the user has a tendency to complain, it is necessary to deal with the user's fault as soon as possible before the user complains. Therefore, when the user has a tendency to complain, the priority of the user is increased to the best.

As shown in Figure 2, according to the dimensions that affect user priority sorting, the historical behavior data of the number of reported faults, the number of confirmed faults, the number of fault-reported calls, the number of complaints, the number of power failure false positives, and the number of calls for this fault are analyzed. , as shown in Table 1, capture the historical fault repair data of four users:

Table 1

Modeling using the underlying data yields the following model:

Failure Accuracy:

x=∑ _t∈1year A _t /∑ _t∈1year B _t

Frequency of incoming calls reported by users:

y=(∑ _t∈1year T _t /∑ _t∈1year A _t ) ²

User false positive probability:

Probability of user complaints:

P(C)=(Σ _t∈1year C _t /Σ _t∈1year T _t )·E

Among them, coefficient A is the total number of reported faults, coefficient B is the number of reported faults confirmed as real faults, coefficient T is the total number of incoming calls for faults, coefficient C is the number of complaints, coefficient D is the number of falsely reported faults, and coefficient E is the number of faults reported for this time. The number of fault calls, ∑ t _{∈ 1year} ρ _t indicates that one year is the statistical time, where t represents the time, and t ∈ 1year indicates that the statistical time is one year.

According to user habits, the priority sorting model determines two parameters, parameter 1 is F(n):

F(n)=x+y+z,

F(n)=∑ _t∈1year (A ³ +BT ² +ABD ² ) _t /∑ _t∈1year (BA ² ) _t ,

Among them, the values of T and D have a greater impact on the parameter F(n), that is, the more the number of historical false alarms, the more frequent repair reports for the same fault, and the lower the ranking of user priorities;

Parameter 2 is the user complaint probability P(C).

According to the user data obtained above, the user priority score and complaint probability are calculated, as shown in Table 2:

Table 2

According to the model calculation, two parameters F(n) and P(C) are obtained to determine the user priority order, as follows:

If the parameter F(n)≤5, the user priority is rated as A;

If parameter 5<F(n)≤10, the user priority is rated as B level;

If the parameter 10<F(n)≤15, the user priority is rated as C level;

If parameter 15<F(n), the user priority is rated as D;

Parameter P(C) is used as the second criterion for user priority evaluation. When parameter P(C)<1, user classification is determined by the result of parameter F(n). When parameter P(C)≥1, directly The user is rated A.

As shown in Table 2, if the priority score of user 2 is greater than 5 points, it should be rated as level B, but since the probability of complaint is greater than 1 and there is a tendency to complain, the user is rated as the best level of level A.

The above disclosed are the preferred embodiments of the present invention, which cannot be used to limit the scope of rights of the present invention. Therefore, equivalent changes made according to the scope of the patent application of the present invention are still within the scope of the present invention.

Claims (3)

1. an active service assistance determination method based on power report failure, is characterized in that, described method comprises the steps: a. Analyze the dimensions that affect user prioritization; b. Capture user historical behavior data; c. Establish a user priority ranking model; d. Through user historical behavior data and user priority sorting model, user priority evaluation is realized; The dimensions of the user prioritization model in step c are as follows: c1. The accuracy of the fault reported by the user; c2. Frequency of incoming calls reported by users for faults; c3. The probability of user complaints; c4. The probability of user false positives; Using the accuracy rate of faults reported by users, the frequency of incoming calls reported by users, the probability of user complaints and the probability of user false alarms, a user priority ranking model is established, as follows: Failure Accuracy: x=Σ _t∈1year A _t /Σ _t∈1year B _t Frequency of incoming calls reported by users: y=(∑ _t∈1year T _t /∑ _t∈1year A _t ) ² User false positive probability:

Probability of user complaints: P(C)=(Σ _t∈1year C _t /Σ _t∈1year T _t )·E Among them, coefficient A is the total number of reported faults, coefficient B is the number of reported faults confirmed as real faults, coefficient T is the total number of incoming calls for faults, coefficient C is the number of complaints, coefficient D is the number of falsely reported faults, and coefficient E is the number of faults reported for this time. The number of faulty calls, ∑ _t∈1year ρ _t means one year is the statistical time, where t represents time, t∈1year means the statistical time is one year; according to user habits, the priority sorting model determines two parameters, parameter 1 is F (n): F(n)=x+y+z, F(n)=∑ _t∈1year (A ³ +BT ² +ABD ² ) _t /∑ _t∈1year (BA ² ) _t , Among them, the values of T and D have a greater impact on the parameter F(n), that is, the more the number of historical false alarms, the more frequent repair reports for the same fault, and the lower the ranking of user priorities; Parameter 2 is the user complaint probability P(C); According to the established user priority ranking model, two parameters F(n) and P(C) are calculated and used to determine the user priority ranking, as follows: If the parameter F(n)≤5, the user priority is rated as A; If parameter 5<F(n)≤10, the user priority is rated as B level; If the parameter 10<F(n)≤15, the user priority is rated as C level; If parameter 15<F(n), the user priority is rated as D; The parameter P(C) is used as the second criterion for user priority evaluation. When the parameter P(C)<1, the user classification is determined by the result of the parameter F(n); when the parameter P(C)≥1, directly The user is rated A. 2. a kind of active service auxiliary determination method based on power report failure as claimed in claim 1, is characterized in that: in step a, the dimension analysis that affects user's prioritization is as follows: a1. The influence of the correct rate of faults reported by users on the priority of users; a2. The impact of user complaints on user priority; a3. The impact of user false positives on user priority; a4. The influence of the number of calls made by the user for the same fault on the priority. 3. a kind of active service auxiliary determination method based on power report failure as claimed in claim 1, is characterized in that: the user historical behavior data that needs to be captured in step b is as follows: b1. Take the year as the unit, capture the behavior data of users reporting power outages within one year of history; b2. Capture the fault data reported by users; b3. Capture the data confirmed as the fault in the fault reported by the user; b4. Capture the total number of calls reported by users for faults; b5. Capture the number of user complaints; b6. Capture the number of user falsely reported faults; b7. Capture the number of calls reported by the user for this fault.

Images