CN112598343A

CN112598343A - Risk inspection method, device, equipment and storage medium for data center

Info

Publication number: CN112598343A
Application number: CN202110248744.5A
Authority: CN
Inventors: 吕鑫; 安柯; 徐蓉; 苏永华
Original assignee: Shanghai Yovole Computer Network Co ltd; Shanghai Youfu Zhishu Yunchuang Digital Technology Co ltd
Current assignee: Shanghai Yovole Computer Network Co ltd; Shanghai Youfu Zhishu Yunchuang Digital Technology Co ltd
Priority date: 2021-03-08
Filing date: 2021-03-08
Publication date: 2021-04-02
Anticipated expiration: 2041-03-08
Also published as: CN112598343B

Abstract

The application relates to a risk inspection method, a risk inspection device, risk inspection equipment and a storage medium for a data center, wherein the method comprises the following steps: acquiring first inspection information determined by identifying information acquired by monitoring equipment in a data center, wherein the first inspection information comprises inspection states of all inspection points; acquiring second inspection information recorded by an inspection worker, wherein the second inspection information comprises inspection description texts and operation data of each inspection point; determining a risk word of each inspection point based on the inspection state and operation data of each inspection point and the inspection description text; determining a first estimation value and a second estimation value of each inspection point; and determining the risk level of the data center by utilizing a pre-constructed risk level space coordinate system according to the first estimation value and the second estimation value of each inspection point. This application can avoid patrolling and examining personnel's omission to examine, reduces the artifical potential safety hazard of patrolling and examining and reexamining and bringing, promotes data center's safe operation efficiency.

Description

Risk inspection method, device, equipment and storage medium for data center

Technical Field

The present application relates to the field of computer technologies, and in particular, to a method, an apparatus, a device, and a storage medium for risk inspection of a data center.

Background

A data center generally refers to a physical space for centralized processing, storage, transmission, exchange and management of information, and computer devices, server devices, network devices, storage devices, and the like are generally regarded as key devices of a network core room, and these key devices need certain environmental factors for operation, such as power supply systems, rack systems, and infrastructure such as monitoring systems.

To ensure proper operation of critical equipment and infrastructure, inspection personnel are required to perform a full amount of inspections at intervals. When in inspection, inspection personnel mainly confirm whether each inspection point has abnormity, such as the temperature of a transformer, the voltage and the current of each cabinet, the state of an indicator light and the like, and then submit the operation data of the inspection point and the text description of an inspection conclusion in a paper form signing mode. But key equipment and infrastructure are distributed in different rooms, and inspection personnel can easily miss inspection. Moreover, the manual inspection mode depends on the subjectivity of inspection personnel, so that the condition of missed inspection exists, and the inspection quality and time can have unmanageable blind spots. In addition, because different inspection personnel describe the inspection results differently, the text descriptions used by the inspection personnel are different, the severity of the potential safety hazard expressed by the text descriptions is also not used, and managers need to distinguish the risk level to be expressed from the text descriptions when the managers manually and periodically check the inspection quality every time, so that a large amount of time and cost are required, the results of the reexamination of different managers are inconsistent, and the safe operation of the data center is seriously influenced.

Disclosure of Invention

The application provides a risk inspection method, device, equipment and storage medium for a data center, and the risk level is determined by combining information recorded by inspection personnel and information collected by monitoring equipment, so that the omission of inspection of the inspection personnel can be avoided, potential safety hazards caused by manual inspection and rechecking are reduced, and the safe operation efficiency of the data center is improved.

In one aspect, the application provides a risk inspection method for a data center, the method including:

acquiring first inspection information determined by identifying information acquired by monitoring equipment in a data center, wherein the first inspection information comprises an inspection state of each inspection point, and the monitoring equipment is set based on the position of each inspection point;

acquiring second inspection information recorded by an inspection worker, wherein the second inspection information comprises an inspection description text and operation data of each inspection point;

determining a risk word of each inspection point based on the inspection state of each inspection point, the inspection description text and the operation data of each inspection point, wherein the risk word is used for evaluating the risk condition of the inspection point;

for each inspection point, determining a first estimation value and a second estimation value of the inspection point according to risk factors of risk words of the inspection point under each preset risk level;

calculating the average value of the first estimation values of all the inspection points to obtain a target first estimation value;

calculating the average value of the second estimation values of all the inspection points to obtain a target second estimation value;

obtaining data point coordinates according to the target first estimation value and the target second estimation value;

and determining a coordinate area of the data point coordinate in a pre-constructed risk level space coordinate system, and determining a preset risk level corresponding to the coordinate area as the risk level of the data center.

Another aspect provides a risk inspection device for a data center, the device including:

the system comprises a monitoring information acquisition module, a data center and a data processing module, wherein the monitoring information acquisition module is used for acquiring first inspection information determined by identifying information acquired by monitoring equipment in the data center, the first inspection information comprises an inspection state of each inspection point, and the monitoring equipment is set based on the position of each inspection point;

the manual record acquisition module is used for acquiring second inspection information recorded by an inspection worker, wherein the second inspection information comprises an inspection description text and operation data of each inspection point;

the risk word determining module is used for determining a risk word of each inspection point based on the inspection state of each inspection point, the inspection description text and the operation data of each inspection point, wherein the risk word is used for evaluating the risk condition of the inspection point;

the target value determining module is used for determining a first estimated value and a second estimated value of each inspection point according to risk factors of the risk words of the inspection points under each preset risk level for each inspection point;

the first estimation value calculation module is used for calculating the average value of the first estimation values of all the inspection points to obtain a target first estimation value;

the second estimated value calculation module is used for calculating the average value of the second estimated values of all the inspection points to obtain a target second estimated value;

the data point coordinate determination module is used for obtaining data point coordinates according to the target first estimation value and the target second estimation value;

and the risk level determination module is used for determining a coordinate area of the data point coordinate in a pre-constructed risk level space coordinate system, and determining a preset risk level corresponding to the coordinate area as the risk level of the data center.

Another aspect provides a computer storage medium having at least one instruction or at least one program stored therein, the at least one instruction or the at least one program being loaded and executed by a processor to implement the risk inspection method for a data center as described above.

According to the method and the device, the risk words of each inspection point are determined by combining the first inspection information acquired by the monitoring equipment based on the positions of the inspection points with the second inspection information recorded by the inspection personnel, so that the inspection omission of the inspection personnel is avoided, and the potential safety hazard caused by the subjective awareness of the inspection personnel to the risk description is reduced; according to the risk factors of each risk word under each preset risk level, the first estimation value and the second estimation value of each inspection point are determined, the risk level of the data center is determined through the risk level space coordinate system which is constructed in advance, inspection information can be matched with the preset risk levels, the inspection quality is guaranteed to be evaluated by using a unified evaluation standard, differences caused by manual review of managers are reduced, and the safe operation efficiency of the data center is improved.

Drawings

In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative efforts.

Fig. 1 is a schematic diagram of an implementation environment provided by an embodiment of the present application.

Fig. 2 is a schematic flowchart of a risk inspection method for a data center according to an embodiment of the present application.

Fig. 3 is a schematic flowchart of another risk inspection method for a data center according to an embodiment of the present application.

Fig. 4 is a schematic flowchart of another risk inspection method for a data center according to an embodiment of the present application.

Fig. 5 is a schematic flowchart of determining a polling state according to an embodiment of the present application.

Fig. 6 is an exemplary diagram of a visualization presentation provided by an embodiment of the present application.

Fig. 7 is a schematic flow chart of determining risk words provided by an embodiment of the present application.

Fig. 8 is a schematic flow chart for determining a first risk word provided by an embodiment of the present application.

Fig. 9 is a schematic flow chart of determining a first estimated value and a second estimated value according to an embodiment of the present application.

Fig. 10 is a schematic flowchart of another risk inspection method for a data center according to an embodiment of the present application.

Fig. 11 is a diagram of an example of a risk level spatial coordinate system provided by an embodiment of the present application.

Fig. 12 is a schematic structural diagram of a risk inspection device for a data center according to an embodiment of the present application.

Fig. 13 is a hardware structural diagram of an apparatus for implementing the method provided by the embodiment of the present application.

Detailed Description

In order to solve the technical problem, an embodiment of the present application provides a risk inspection method for a data center. To make the objects, technical solutions and advantages of the present application more clear, the present application will be further described in detail with reference to the accompanying drawings. It is to be understood that the described embodiments are merely a few embodiments of the present application and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.

It should be noted that the terms "first," "second," and the like in the description and claims of this application and in the drawings described above are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used is interchangeable under appropriate circumstances such that the embodiments of the application described herein are capable of operation in sequences other than those illustrated or described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or server that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.

Referring to fig. 1, a schematic diagram of an implementation environment provided by an embodiment of the present application is shown, and as shown in fig. 1, the implementation environment may include at least a data center 10 and a server 20.

The data center 10 is provided with key devices of a network core machine room, such as computer devices, server devices, network devices, and storage devices, and a power supply system, a rack system, and a monitoring system, which are used to provide infrastructure required for the operation of the key devices. It is understood that other critical equipment and infrastructure may be located within the data center 10, and this description is not intended to be limiting.

The server 20 may comprise a server operating independently, or a distributed server, or a server cluster consisting of a plurality of servers, and the server 20 may further comprise a network communication unit, a processor, a memory, and the like. Specifically, the server 20 may acquire the automatically recognized information and the information uploaded by the inspection personnel through an infrastructure or the like, and then determine the risk level of the data center based on the information.

For convenience of description, a risk inspection method for a data center according to an embodiment of the present application is described below with a server as an execution subject. Fig. 2 is a schematic flow chart of a risk inspection method for a data center according to an embodiment of the present application, and the present specification provides the method operation steps according to the embodiment or the flowchart, but more or less operation steps may be included based on conventional or non-inventive labor. The order of steps recited in the embodiments is merely one manner of performing the steps in a multitude of orders and does not represent the only order of execution. In practice, the system or server product may be implemented in a sequential or parallel manner (e.g., parallel processor or multi-threaded environment) according to the embodiments or methods shown in the figures. Specifically, as shown in fig. 2, the method may include:

s210, first patrol inspection information determined by identifying the information collected by the monitoring equipment in the data center is obtained, wherein the first patrol inspection information comprises patrol inspection states of all patrol inspection points, and the monitoring equipment is set based on the positions of all patrol inspection points.

In the embodiment of the application, the inspection point refers to the key point that needs to be inspected, and every inspection point corresponds an inspection object, and whether the inspection object corresponding to this inspection point is inspected in the preset inspection period is inspected by the inspection state characterization, and the inspection state of every inspection point can include that inspection has been performed, the generation is inspected and inspection is not performed. The inspection object is inspected by the inspection point corresponding target inspection personnel, and the inspection object is not inspected by the inspection point corresponding target inspection personnel.

The monitoring device refers to a device with a video recording function or a face recognition function, such as a camera or a video recorder, and the monitoring device in the data center may include one or more devices. In specific implementation, one monitoring device may be set for each inspection point, or one monitoring device may be set for a plurality of inspection points. The determination of the patrol point can be manual determination, or can be automatic determination by the server through dynamic identification according to the environment image in the data center.

Based on this, in a possible embodiment, as shown in fig. 3, before the step S210 is implemented, the method may further include:

s201, acquiring an environment image in the data center, and performing image recognition on the environment image to recognize the type of the target object in the environment image.

The environment image represents environment information in the data center, and the environment image comprises the positions of the routing inspection objects. The server can obtain the environment image from a preset database and also can obtain the environment image from a monitoring system. The target object may be a cabinet, a transformer, a computer, etc. in the data center, and is not limited herein.

The server may identify the environmental image using a target image identification algorithm to identify a type of the target object in the environmental image. The target image recognition algorithm may be a convolutional neural network-based target image recognition algorithm or a support vector machine-based target image recognition algorithm, and the like, and this specification is not limited in particular.

S202, if the type is in the preset type list, determining the target object as the inspection object, and recording the position information of the inspection object.

In this embodiment, the preset type list stores all types of objects to be inspected, such as cabinets and transformers. If the type of the identified target object is in the preset type list, the target object can be considered as an object needing to be inspected, namely an inspection object, and then the position information of the target object output by the target image recognition algorithm is recorded.

S203, determining the positions of all inspection points in the data center according to the position information of the inspection objects, wherein the inspection points correspond to the inspection objects one to one.

And for each inspection object, determining the vacant position of the environment around the inspection object based on the position information of the inspection object, and then selecting a target position from the vacant positions according to a preset rule as the position of the inspection point corresponding to the inspection object. The preset rule preferentially selects the inspection object entrance as a target position, so that the human face characteristics and the behavior characteristics of inspection personnel can be captured conveniently.

And S204, setting a monitoring device at the position of each inspection point.

Because the problem of missed-check exists, a certain inspection object is easily omitted when the inspection personnel inspect, and the inspection time is uncontrollable due to the personnel management problem. The monitoring equipment is arranged at the position of each inspection point, inspection personnel need to check in through the monitoring equipment when inspecting each inspection object, the server can determine whether the inspection personnel inspect the time frame or not based on the check-in information acquired by the monitoring equipment, and can determine whether the inspection personnel inspect the time frame or not.

The setting of the inspection period can be preset in the monitoring equipment, and the monitoring equipment allows inspection personnel to check in only in the inspection period; the monitoring device can allow the inspection personnel to check in at any time period, and whether the server judges the check-in time period or not can be pre-stored by the server. The mode of judgment by the server requires the server to store the inspection time period corresponding to each monitoring device, and when the data center is large enough, the number of objects to be inspected is increased gradually, and the number of monitoring devices is increased accordingly. With the increase of the monitoring equipment, the inspection period to be maintained also increases, resulting in the increase of the maintenance cost.

Based on the above description, in one possible implementation, as shown in fig. 4, after step S204 is implemented, the method may further include:

s205, setting at least one inspection time interval for each monitoring device, and setting target inspection personnel corresponding to each inspection time interval, wherein the inspection time intervals are not overlapped.

In practical applications, the patrol personnel may perform the patrol operation more than once a day, and thus one or more patrol periods may be set for each monitoring device. The same patrol personnel can patrol in each patrol period, and different patrol personnel can patrol. Different patrol personnel have different credibility, for example, the credibility of the patrol information recorded by the newly on duty patrol personnel is usually less than the credibility of the patrol information recorded by the mature patrol personnel. Therefore, in order to manage and assess the inspection personnel and determine the risk degree according to the credibility of the inspection personnel, the corresponding target inspection personnel can be set for each inspection time period.

And S206, in each inspection period, acquiring the monitoring video stream through the monitoring equipment, and determining the inspection state of the inspection point corresponding to the monitoring equipment according to the monitoring video stream and the target inspection personnel corresponding to the inspection period.

Starting a collection mode by the monitoring equipment at the initial time of each inspection period to start video collection, and transmitting the monitoring video stream collected in the inspection period to the server by the monitoring equipment when the inspection period is terminated; the server compares the check-in personnel in the monitoring video stream with the target check-in personnel to determine the check-in state of the check-out point. And at the termination time of each inspection period, the monitoring equipment also closes the acquisition mode, and after the acquisition mode is closed, the monitoring equipment still acquires the monitoring video stream, but the acquired monitoring video stream is not transmitted to the server and is only stored in the local memory of the monitoring equipment for backup. In specific implementation, the monitoring equipment can transmit the collected monitoring video stream to the server after the check-in personnel finish checking in.

As shown in fig. 5, in a possible implementation manner, the determining, according to the monitored video stream and the target inspection person corresponding to the inspection time period in step S206, the inspection state of the inspection point corresponding to the monitoring device may include, in specific implementation:

s2061, extracting the human face features of the check-in patrollers from the monitoring video stream.

Specifically, the server can perform face detection on each frame of image in the monitoring video stream so as to screen out a face image to be recognized, which at least comprises one face; and then extracting the face features of the patroller from each face image to be recognized. The extraction method of the face features of the face image to be recognized may be various, and the application does not specifically limit this. For example, the face features can be characterized by converting the face image to be recognized into a set of float32 values through a deep neural network.

S2062, if the human face features of the check-in patrol personnel are not extracted, setting the patrol state of the patrol point corresponding to the monitoring equipment as no patrol.

If the human face features of the patrolling personnel are not extracted, the patrolling personnel are not checked in the patrolling period, namely the patrolling personnel miss the detection of the patrolling object corresponding to the monitoring equipment, and therefore the patrolling state of the patrolling point is set to be not patrolled.

S2063, if the face features of the check-in patrol personnel are extracted, the face features of the target patrol personnel are obtained.

If the human face features of the check-in patrol personnel are extracted, the check-in patrol personnel check-in the patrol period, but whether the check-in patrol personnel are the target patrol personnel needs to be checked. The server can directly acquire the face features of the target inspection personnel from a preset database, and the preset database stores the face features of all the inspection personnel.

S2064, detecting whether the face characteristics of the check-in patrol personnel are consistent with the face characteristics of the target patrol personnel.

If yes, go to step S2065; if not, go to step S2066.

S2065, setting the inspection state of the inspection point corresponding to the monitoring equipment as inspected.

S2066, the inspection state of the inspection point corresponding to the monitoring equipment is set as the substitute inspection.

When the inspection state of each monitoring device corresponding to the inspection point is determined to be completed, the monitoring indicator lamp of the monitoring device can be changed according to the inspection state, different inspection states correspond to different colors of the indicator lamps, a manager can check whether the monitoring system determines whether to leak or not according to different colors of the indicator lamps or inspection states, for example, a red indicator lamp can be used when the inspection is not performed, a yellow indicator lamp can be used when the inspection is performed instead, and a green indicator lamp can be used when the inspection is performed. Every object department of patrolling and examining all is provided with supervisory equipment, in case the personnel of patrolling and examining patrolled and examined the object and had patrolled and examined or checked in at this point of patrolling and examining in the period, all will be recorded by supervisory equipment to show with different pilot lamp colours.

For the server, in order to enable the manager to master the inspection information in real time, after the inspection of the inspection personnel is completed each time, namely after the inspection state of the inspection point corresponding to each monitoring device is determined, the server can visually display the inspection state of the inspection point corresponding to each monitoring device. As shown in fig. 6, which is an example of a visualization presentation. In fig. 6, the inspection period of each inspection object can be clearly checked, and whether each inspection period has completed inspection or not, the inspection points passed by the inspection personnel in the inspection period are faithfully recorded by the monitoring equipment and fed back in the monitoring system, for example, the inspection object 4 in fig. 6 has two inspection periods, and the inspection period is marked with a color different from the inspection period due to the fact that inspection is not completed in the first inspection period.

In some embodiments, in order to avoid repeated processing of the video stream, when receiving the monitoring video stream collected by the monitoring device, that is, before implementing step S2061, the server may first obtain the inspection state of the inspection point corresponding to the monitoring device, and if the inspection state is inspection or inspection substitute, the server may not process the monitoring video stream; if the polling status is not polling or empty (i.e., not acquired), step S2061 is performed.

The face recognition is carried out according to the monitoring video stream, so that the face features of the check-in patrol personnel can be effectively extracted, and the problem of missing detection is avoided; then the human face characteristics of the check-in personnel are compared with the human face characteristics of the target check-in personnel, so that the check-in time and the check-in personnel can be effectively managed, and blind spots in management are solved.

And S220, acquiring second inspection information recorded by the inspection personnel, wherein the second inspection information comprises an inspection description text and operation data of each inspection point.

After the inspection is finished, the inspection personnel records the inspection information related to the inspection, and the operation data of each inspection point refers to the index value of the inspection point, such as a current value, a voltage value or a temperature and the like; the patrol description text is usually a description of the safety risk of each or some patrol points or corresponding patrol objects, for example, "current is normal, voltage is slightly higher, and cabinet indicator light is dim". It can be understood that the index of each inspection point is related to the inspection object corresponding to the inspection point, and is not specifically limited herein. As shown in table 1, which is one example of the operation data of the patrol point.

TABLE 1

And S230, determining the risk words of each inspection point based on the inspection state, the inspection description text and the operation data of each inspection point.

In the embodiment of the application, the risk words are used for indicating the risk conditions of the inspection points, such as "higher", "normal", "lower", "abnormal", and the like, and the risk words of each inspection point can be generally obtained directly from the inspection description text, but the inspection description text may only cover a part of information of the inspection points, and the inspection description text has subjectivity. To avoid missed inspection by the inspection personnel and such subjectivity, in a possible embodiment, as shown in fig. 7, the step S230 may include:

s2301, determining a first risk word of each inspection point based on the inspection state and the operation data of each inspection point.

The inspection state represents whether the inspection point corresponds to the inspection object or not, and whether the filled operation data are credible or not can be determined by combining the inspection state, so that the first risk words are determined.

In one possible implementation, as shown in fig. 8, the step S2301 may include, in specific implementation:

s23011, for each inspection point, if the inspection state of the inspection point is not inspection, determining a preset risk word corresponding to the inspection point as a first risk word of the inspection point.

Wherein the preset risk words represent default risk conditions. The preset database stores preset risk words corresponding to each inspection point, and the preset risk words corresponding to each inspection point can be the same or different and are not specifically limited. For example, for the cabinet 1, since there is no device in the data center that depends on its operation, if it is not checked, the preset risk word corresponding to the cabinet 1 may be set to "abnormal"; for the cabinet 2, because there are many devices depending on the operation of the data center, if the devices are not checked, the preset risk word corresponding to the cabinet 2 may be set to "extreme abnormal".

In specific implementation, the server can also acquire whether alarm information exists in the inspection point under the condition that the inspection state of the inspection point is not inspected; if the first risk word exists, the first risk word can be determined according to the alarm information; if not, the preset risk words corresponding to the inspection point can be determined as the first risk words of the inspection point. For determining the first risk word according to the alarm information, the server may determine according to the importance degree of the alarm information or the type of the alarm information, and the alarm information with different importance degrees or different types has different first risk words, which is not specifically limited herein. For example, an alarm message with a higher hierarchical level of importance corresponds to a first risk term of high risk.

S23012, if the inspection state of the inspection point is inspection or inspection generation, comparing the operation data of the inspection point with the statistical indexes of the inspection point to obtain a first risk word of the inspection point.

In the embodiment of the application, the statistical indexes are determined based on historical inspection data of the inspection point, the statistical indexes comprise an average value of each index item, and the index item refers to an index, such as voltage, current or temperature, for detecting the inspection point corresponding to the inspection point and the inspection object. Specifically, the server acquires historical patrol data of the patrol point, and performs statistical analysis on the historical patrol data to obtain statistical indexes; for each index item in the operation data, calculating the difference between the index value of the index item and the average value of the index item in the statistical index to obtain an index difference value; and searching a risk word corresponding to the index difference in a preset index item table to serve as a first risk word of the inspection point. That is, for each index, one first risk term will be generated, and thus the first risk term for the patrol point may include one or more.

And the preset index items are recorded with risk words of each index item in each patrol point in different index difference ranges. As shown in table 2, it is an example of the preset index item table. As can be seen from table 2, once the index difference range is determined, the corresponding risk word can be quickly obtained.

TABLE 2

And S2302, performing text recognition on the patrol inspection description text, and determining a second risk word of the patrol inspection point corresponding to the patrol inspection description text.

In practical application, the patrol inspection description text is expressed differently by each patrol inspection person, and if the contents to be expressed can be identified from the patrol inspection description text recorded by the patrol inspection person, the accuracy of determining the risk level can be increased. The method comprises the steps of carrying out state recognition on the patrol inspection description text by adopting an attribute-level emotion classification method based on a dictionary to determine a second risk word corresponding to a patrol inspection point of the patrol inspection description text, and mainly comprising two steps of marking emotion expression and processing emotion conversion words.

The marked emotion expressions are mainly used for finding out each emotion expression in a sentence of the patrol description text and judging the emotion tendency corresponding to each emotion expression, and each emotion expression may contain one or more attributes. For example, for the sentence "current is normal, voltage is slightly higher, cabinet indicator light is dim", "normal", "slightly higher", and "dim" are all labeled attributes.

It will be appreciated that each inspector may have a number of different representations of the same risk, and thus when marking emotional expressions, a number of similar attributes may be marked which represent the same type of risk. And thus similar attributes can often be expressed in terms of the same class of risk terms. In specific implementation, when the server marks emotional expressions, the routing inspection description texts can be automatically classified by adopting a preset algorithm, and similar attributes are classified into risk words. The preset algorithm may be, for example, a support vector machine algorithm, a decision tree algorithm, or a naive bayes algorithm, which is not specifically limited in this application.

The emotion conversion words refer to words or phrases capable of changing emotional tendency, the emotional conversion times can include various types, such as common negative word types or turning word types of "not", "but", and the like, the viewpoints before and after the words appear usually have opposite emotional tendency, and in this case, not only the viewpoint tendency of one side but also the viewpoint tendency of the other side need to be considered, and thus special treatment is needed. For example, in the above-mentioned case "the voltage is slightly higher", it may become "the voltage is normal but is already at the upper limit of the critical voltage" when another inspection person makes an expression, and in this case, in addition to the mark "normal", a mark "is required, but" the following conversion words affecting the normal emotion, that is, "critical" and "upper limit" are characterized.

S2303, the first risk words and the second risk words are fused, and risk words of each inspection point are obtained.

After emotion expression is marked and emotion conversion words are processed, the server can obtain second risk words of the recorded index items of the inspection points, then the first risk words of each inspection point are fused, the second risk words are added into the first risk words, and the risk words of each inspection point are obtained. For example, for inspection point a, the corresponding first risk word is "normal" (current value), and the corresponding second risk word is "dim" (indicator light), and the risk words of inspection point a are "normal" (current value) and "dim" (indicator light).

S240, for each inspection point, determining a first estimation value and a second estimation value of the inspection point according to risk factors of the risk words of the inspection point under each preset risk level.

In the embodiment of the application, words expressing risk levels defined in advance, namely preset risk levels, are set in the preset risk lexicon, and the preset risk lexicon mainly comprises 8 categories of low risk, medium risk, high risk, ultrahigh risk, no risk, potential medium risk, potential high risk, potential ultrahigh risk and the like. Wherein, potentially medium risk means the possibility of having potentially medium risk, potentially high risk means the possibility of having potentially high risk, and potentially ultra high risk means the possibility of having potentially ultra high risk.

The first estimate represents an estimate of the likelihood of the risk level and the second estimate represents an estimate of the degree of activation of the risk level, the higher the first and second estimates, the higher the risk level and thus the higher the degree of risk. The risk words in each inspection point can comprise one or more, and each risk word has a first estimation value and a second estimation value, so that the overall score of the inspection point can be calculated. The server can determine a first estimation value and a second estimation value of the inspection point according to a first risk factor of each risk word under each preset risk level, a second risk factor of each risk word under each preset risk level and the risk importance of the inspection point.

Specifically, in one possible implementation, as shown in fig. 9, the step S240 may include, in implementation:

s2401, for each inspection point, acquiring the risk importance of the inspection point.

In the embodiment of the present application, the risk importance is a specific value, which represents the degree of risk importance, and is an example of the risk importance of the patrol point as shown in the following table. It is understood that the risk importance of each patrol point in table 3 is only an example, and the setting of the risk importance is not particularly limited.

TABLE 3

S2402, determining a first estimated value of the inspection point according to the risk importance of the inspection point and first risk factors of risk words of the inspection point under each preset risk level.

The first risk factors represent the possibility of the word under the preset risk level, and each risk word has a corresponding first risk factor under different preset risk levels. Specifically, for each preset risk level, the server may determine the product of the risk importance of the inspection point at the preset risk level and the first risk factor of each risk word at the preset risk level as the first risk coefficient of the inspection point at the preset risk level; and then selecting the maximum value of the first risk coefficients of the inspection points under each preset risk level to obtain a first estimated value of the inspection point. As shown in table 4, which is an example of a first risk factor for each risk term at different preset risk levels.

TABLE 4

Specifically, a set of risk level spaces is set

It contains 8 preset risk levels

Indicating inspection points

Risk word set of

，

Indicating inspection points

To (1)

The individual risk word or words are,

to represent

At a predetermined risk level

The first risk factor of. If each risk word can independently contribute to the overall risk level, the inspection point

At the risk level

（

) First risk factor of

Can be given by the following equation:

wherein the content of the first and second substances,nto patrol the number of risk terms in point j,

to a patrol point

Risk importance of. Then, the first estimated value of the patrol point j is, each

Maximum value of (2).

S2403, determining a second estimated value of the inspection point according to the inspection point risk importance and second risk factors of the risk words of the inspection point under each preset risk level.

The second risk factors represent the activation degree of the word under the preset risk level, and each risk word has a corresponding second risk factor under different preset risk levels. Specifically, for each preset risk level, the server may determine the product of the risk importance of the inspection point at the preset risk level and the second risk factor of each risk word at the preset risk level as the second risk coefficient of the inspection point at the preset risk level; and then selecting the maximum value of the second risk coefficients of the inspection points under each preset risk level to obtain a second estimated value of the inspection points. As shown in table 5, which is an example of a second risk factor for each risk term at different preset risk levels.

TABLE 5

Specifically, a set of risk level spaces is set

It contains 8 preset risk levels

Indicating inspection points

Risk word set of

，

Indicating inspection points

To (1)

The individual risk word or words are,

to represent

At a predetermined risk level

The second risk factor of the next inspection point

At the risk level

（

) Second risk factor of

Can be given by the following equation:

to a patrol point

Risk importance of. Then, the second estimated value of the patrol point j is, i.e., each

Maximum value of (2).

It should be noted that, in the above embodiment, the preset risk terms corresponding to the inspection point, the risk importance of the inspection point at each preset risk level, and the preset values of the first risk factor and the second risk factor of each risk term at each preset risk level may be obtained by querying an experience table in the database. Each preset value stored in the experience table is obtained by extracting the server based on experience data of the data center in advance, for example, the server can obtain historical patrol data, and then model training is performed on a preset neural network model by using the historical patrol data as a training sample. In addition, the risk importance is a value greater than 1, and the first risk factor and the second risk factor are both values between 0 and 1.

And S250, calculating the average value of the first estimation values of all the inspection points to obtain a target first estimation value.

For example, suppose that the data center has 4 polling points, and the first estimated values corresponding to the 4 polling points are: 5.08, 6.18, 7.55 and 7.6, the target first estimate is the average of these 4 first estimates, i.e. 6.60.

And S260, calculating the average value of the second estimation values of all the inspection points to obtain a target second estimation value.

For example, for the above 4 polling points, the corresponding second estimated values are 3.65, 4.28, 5.04, and 6.17, respectively, and then the target second estimated value is the tie value of the 4 second estimated values, i.e., 4.78.

And S270, obtaining data point coordinates according to the first target estimation value and the second target estimation value.

Specifically, the data point coordinates can be obtained by using the first target estimation value as an abscissa value of the risk level space coordinate system and the second target estimation value as an ordinate value of the risk level space coordinate system. For example, for the 4 patrol points described in steps S250 and S260, the resulting data point coordinates may be represented as (6.60, 4.78).

And S280, determining a coordinate area of the data point coordinate in the risk level space coordinate system, and determining a preset risk level corresponding to the coordinate area as the risk level of the data center.

In the embodiment of the application, the server respectively corresponds 8 preset risk levels to one coordinate area in a risk level space coordinate system to form a two-dimensional area of the whole risk level space coordinate system.

As shown in fig. 10, in a possible implementation, before the step S250 is implemented, the method may further include:

s248, establishing a risk level space coordinate system, wherein the abscissa of the risk level space coordinate system represents a first estimation value, and the ordinate of the risk level space coordinate system represents a second estimation value;

and S249, carrying out region division on the risk level space coordinate system according to the shape of the Chinese character 'mi', and obtaining a coordinate region corresponding to each preset risk level.

The server can obtain the maximum value of the abscissa and the maximum value of the ordinate, determine a target area according to the maximum value of the abscissa and the maximum value of the ordinate, and then perform area division on the target area according to the shape of a Chinese character 'mi', so as to obtain a coordinate area corresponding to each preset risk level. The maximum value of the abscissa represents the maximum value of the first estimation value, the maximum value of the ordinate represents the maximum value of the second estimation value, the target area can be a rectangle or a square area, and the shape of the area depends on the setting of the maximum value of the ordinate and the maximum value of the abscissa.

Specifically, if a represents the maximum value of the abscissa, b represents the maximum value of the ordinate, x represents the abscissa, and y represents the ordinate, then the region surrounded by y = b, x = a, and x =0 and y =0 can be determined as the target region; then, the target area is cut into 8 coordinate areas by y = (b/a) x, y = b/2, x = a/2 and y = - (b/a) x + b straight lines, each coordinate area corresponding to a preset risk level. It is understood that a and b are the same, b/a =1, and- (b/a) = -1.

As shown in fig. 11, which is an example of a risk level space coordinate system, in fig. 11, a target area is a square, and as can be seen from the figure, a server divides 8 preset risk levels into an explicit risk level and a potential risk level, where the explicit risk level includes: low risk, medium risk, high risk and ultra high risk, potential risk classes include: no risk, potentially medium risk, potentially high risk and potentially ultra high risk. Generally, the risk levels of the data center are all in a potential risk level, and only when the first estimation value exceeds a certain value, the risk level is changed into a definite risk level.

Based on the same inventive concept as the method embodiment, an embodiment of the present application further provides a risk inspection apparatus for a data center, as shown in fig. 12, the apparatus 1300 may include:

a monitoring information obtaining module 1310, configured to obtain first inspection information determined by identifying information collected by monitoring equipment in a data center, where the first inspection information includes an inspection state of each inspection point, and the monitoring equipment is set based on the position of each inspection point;

a manual record obtaining module 1320, configured to obtain second inspection information recorded by an inspection worker, where the second inspection information includes an inspection description text and operation data of each inspection point;

a risk word determining module 1330, configured to determine a risk word for each inspection point based on the inspection state of each inspection point, the inspection description text, and the operation data of each inspection point, where the risk word is used to evaluate the risk condition of the inspection point;

the target value determining module 1340 is configured to determine, for each inspection point, a first estimated value and a second estimated value of the inspection point according to the risk factors of the risk words of the inspection point at each preset risk level;

a first estimated value calculating module 1350, configured to calculate an average value of the first estimated values of the inspection points to obtain a target first estimated value;

a second estimated value calculating module 1360, configured to calculate an average value of second estimated values of each inspection point, to obtain a target second estimated value;

a data point coordinate determination module 1370, configured to obtain a data point coordinate according to the target first estimation value and the target second estimation value;

the risk level determining module 1380 is configured to determine a coordinate region of the data point coordinate in a pre-constructed risk level spatial coordinate system, and determine a preset risk level corresponding to the coordinate region as a risk level of the data center.

In one possible implementation, the apparatus 1300 may further include:

the environment image acquisition module is used for acquiring an environment image in the data center and carrying out image recognition on the environment image so as to recognize the type of a target object in the environment image;

the inspection object determining module is used for determining the target object as the inspection object and recording the position information of the inspection object when the type is in the preset type list;

the inspection point determining module is used for determining the positions of all inspection points in the data center according to the position information of the inspection object, wherein the inspection points correspond to the inspection object one by one;

and the monitoring equipment setting module is used for setting monitoring equipment at the position of each inspection point.

In one possible implementation, the apparatus 1300 may further include:

the system comprises a polling time interval setting module, a polling time interval setting module and a polling time interval setting module, wherein the polling time interval setting module is used for setting at least one polling time interval for each monitoring device and setting target polling personnel corresponding to each polling time interval, and the polling time intervals are not overlapped with each other;

and the inspection state setting module is used for acquiring the monitoring video stream through the monitoring equipment in each inspection time period, and determining the inspection state of the inspection point corresponding to the monitoring equipment according to the monitoring video stream and the target inspection personnel corresponding to the inspection time period.

In one possible embodiment, the patrol status setting module may include:

the face recognition unit is used for extracting the face features of the check-in patrollers from the monitoring video stream;

the first marking unit is used for setting the inspection state of the inspection point corresponding to the monitoring equipment as not to be inspected under the condition that the human face characteristics of the inspection personnel are extracted;

the target characteristic acquisition unit is used for acquiring the human face characteristics of the target inspection personnel under the condition of extracting the human face characteristics of the check-in inspection personnel;

the characteristic comparison unit is used for detecting whether the human face characteristics of the check-in patrol personnel are consistent with the human face characteristics of the target patrol personnel;

the second marking unit is used for setting the inspection state of the inspection point corresponding to the monitoring equipment as inspected under the condition that the human face characteristics of the checked-in inspection personnel are consistent with the human face characteristics of the target inspection personnel;

and the third marking unit is used for setting the inspection state of the inspection point corresponding to the monitoring equipment as a substitute inspection under the condition that the face characteristics of the inspection personnel who check in are inconsistent with the face characteristics of the target inspection personnel.

In one possible implementation, the risk term determination module 1330 may include:

the first risk word determining unit is used for determining the first risk word of each inspection point based on the inspection state and the operation data of each inspection point;

the second risk word determining unit is used for performing text recognition on the inspection description text and determining a second risk word of the inspection point corresponding to the inspection description text;

and the risk word fusion unit is used for fusing the first risk word and the second risk word to obtain the risk word of each inspection point.

In one possible embodiment, the first risk word determination unit may include:

the inspection state determining unit is used for determining a preset risk word corresponding to the inspection point as a first risk word of the inspection point under the condition that the inspection state of the inspection point is not inspected;

and the data comparison unit is used for comparing the running data of the inspection point with the statistical indexes of the inspection point to obtain a first risk word of the inspection point under the condition that the inspection state of the inspection point is inspection or is a substitute inspection state.

In one possible implementation, the target value determination module 1340 may include:

the importance acquiring unit is used for acquiring the risk importance of each inspection point;

the first estimation value determining unit is used for determining a first estimation value of the inspection point according to the risk importance of the inspection point and first risk factors of risk words of the inspection point under each preset risk level;

and the second estimation value determining unit is used for determining a second estimation value of the inspection point according to the risk importance of the inspection point and second risk factors of the risk words of the inspection point under each preset risk level.

In one possible implementation, the apparatus 1300 may further include:

the coordinate system establishing module is used for establishing a risk level space coordinate system, wherein the abscissa of the risk level space coordinate system represents a first estimation value, and the ordinate of the risk level space coordinate system represents a second estimation value;

and the risk level mapping module is used for carrying out region division on the risk level space coordinate system according to the shape of the Chinese character mi to obtain a coordinate region corresponding to each preset risk level.

It should be noted that, when the apparatus provided in the foregoing embodiment implements the functions thereof, only the division of the functional modules is illustrated, and in practical applications, the functions may be distributed by different functional modules according to needs, that is, the internal structure of the apparatus may be divided into different functional modules to implement all or part of the functions described above. In addition, the apparatus and method embodiments provided by the above embodiments belong to the same concept, and specific implementation processes thereof are described in the method embodiments for details, which are not described herein again.

The embodiment of the application further provides electronic equipment, which comprises a processor and a memory, wherein at least one instruction or at least one program is stored in the memory, and the at least one instruction or at least one program is loaded by the processor and executes the risk inspection method for the data center provided by the method embodiment.

Further, fig. 13 shows a hardware structure diagram of an apparatus for implementing the method provided in the embodiment of the present application, and the apparatus may participate in constituting or containing the device or system provided in the embodiment of the present application. As shown in fig. 13, the device 14 may include one or more (shown here as 1402a, 1402b, … …, 1402 n) processors 1402 (the processors 1402 may include, but are not limited to, a processing device such as a microprocessor MCU or a programmable logic device FPGA, etc.), a memory 1404 for storing data, and a transmission device 1406 for communication functions. Besides, the method can also comprise the following steps: a display, an input/output interface (I/O interface), a Universal Serial Bus (USB) port (which may be included as one of the ports of the I/O interface), a network interface, a power source, and/or a camera. It will be understood by those skilled in the art that the structure shown in fig. 13 is only an illustration and is not intended to limit the structure of the electronic device. For example, device 14 may also include more or fewer components than shown in FIG. 13, or have a different configuration than shown in FIG. 13.

It should be noted that the one or more processors 1402 and/or other data processing circuitry described above may be referred to generally herein as "data processing circuitry". The data processing circuitry may be embodied in whole or in part in software, hardware, firmware, or any combination thereof. Further, the data processing circuitry may be a single, stand-alone processing module, or incorporated in whole or in part into any of the other elements in the device 14 (or mobile device). As referred to in the embodiments of the application, the data processing circuit acts as a processor control (e.g. selection of a variable resistance termination path connected to the interface).

The memory 1404 may be used to store software programs and modules of application software, such as program instructions/data storage devices corresponding to the methods described in the embodiments of the present application, and the processor 1402 executes various functional applications and data processing by running the software programs and modules stored in the memory 1404, so as to implement the risk inspection method for a data center described above. The memory 1404 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 1404 may further include memory remotely located from the processor 1402, which may be connected to the device 14 via a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.

The transmitting device 1406 is used for receiving or sending data via a network. Specific examples of such networks may include wireless networks provided by the communication provider of the device 14. In one example, the transmission device 1406 includes a network adapter (NIC) that can be connected to other network devices through a base station so as to communicate with the internet. In one example, the transmitting device 1406 can be a Radio Frequency (RF) module, which is used for communicating with the internet in a wireless manner.

The display may be, for example, a touch screen type Liquid Crystal Display (LCD) that may enable a user to interact with a user interface of the device 14 (or mobile device).

The embodiment of the application also provides a computer storage medium, where at least one instruction or at least one program is stored in the computer storage medium, and the at least one instruction or the at least one program is loaded and executed by a processor to implement the risk inspection method for a data center provided by the above method embodiment.

Alternatively, in this embodiment, the computer storage medium may be located on at least one of a plurality of network servers of a computer network. Optionally, in this embodiment, the storage medium may include, but is not limited to: a U-disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a removable hard disk, a magnetic or optical disk, and other various media capable of storing program codes.

According to the technical scheme provided by the method embodiment, the risk words of each inspection point are determined by combining the first inspection information acquired by the monitoring equipment arranged based on the position of each inspection point with the second inspection information recorded by the inspection personnel, so that the omission of inspection by the inspection personnel is avoided, and the potential safety hazard caused by the subjective awareness of the inspection personnel on the risk description is reduced; according to the risk factors of each risk word under each preset risk level, the first estimation value and the second estimation value of each inspection point are determined, the risk level of the data center is determined through the risk level space coordinate system which is constructed in advance, inspection information can be matched with the preset risk levels, the inspection quality is guaranteed to be evaluated by using a unified evaluation standard, differences caused by manual review of managers are reduced, and the safe operation efficiency of the data center is improved.

It should be noted that the order of the above embodiments of the present application is only for description, and does not represent the merits of the embodiments. And specific embodiments thereof have been described above. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims may be performed in a different order than in the embodiments and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing may also be possible or may be advantageous.

The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, for the device and electronic apparatus embodiments, since they are substantially similar to the method embodiments, the description is relatively simple, and reference may be made to some descriptions of the method embodiments for relevant points.

The foregoing description has disclosed fully embodiments of the present application. It should be noted that those skilled in the art can make modifications to the embodiments of the present application without departing from the scope of the claims of the present application. Accordingly, the scope of the claims of the present application is not to be limited to the particular embodiments described above.

Claims

1. A risk inspection method for a data center is characterized by comprising the following steps:

2. The method of claim 1, wherein prior to the obtaining the first routing inspection information determined to identify information collected by monitoring equipment in the data center, the method further comprises:

acquiring an environment image in the data center, and performing image recognition on the environment image to recognize the type of a target object in the environment image;

if the type is in a preset type list, determining the target object as an inspection object, and recording the position information of the inspection object;

determining the positions of the inspection points in the data center according to the position information of the inspection object, wherein the inspection points correspond to the inspection object one by one;

and arranging one monitoring device at the position of each inspection point.

3. The method of claim 2, wherein after said disposing one of said monitoring devices at each of said patrol points, said method further comprises:

setting at least one inspection time period for each monitoring device, and setting target inspection personnel corresponding to each inspection time period, wherein the inspection time periods are not overlapped with each other;

and in each inspection period, acquiring a monitoring video stream through the monitoring equipment, and determining the inspection state of the inspection point corresponding to the monitoring equipment according to the monitoring video stream and the target inspection personnel corresponding to the inspection period.

4. The method according to claim 3, wherein the determining the inspection state of the inspection point corresponding to the monitoring equipment according to the monitoring video stream and the target inspection personnel corresponding to the inspection time period comprises:

extracting the human face features of the check-in patrol personnel from the monitoring video stream;

if the human face features of the check-in patrol personnel are not extracted, setting the patrol state of the patrol point corresponding to the monitoring equipment as no patrol;

if the face features of the check-in patrol personnel are extracted, the face features of the target patrol personnel are obtained;

detecting whether the face features of the check-in patrol personnel are consistent with the face features of the target patrol personnel;

if the data are consistent, setting the inspection state of the inspection point corresponding to the monitoring equipment as inspected;

and if the data is inconsistent with the data, setting the inspection state of the inspection point corresponding to the monitoring equipment as the substitute inspection.

5. The method according to claim 1, wherein the determining risk words for each inspection point based on the inspection state of each inspection point, the inspection description text and the operation data of each inspection point comprises:

determining a first risk word of each inspection point based on the inspection state and the operation data of each inspection point;

performing text recognition on the inspection description text, and determining a second risk word of an inspection point corresponding to the inspection description text;

and fusing the first risk words and the second risk words to obtain the risk words of each inspection point.

6. The method according to claim 5, wherein the determining the first risk term for each inspection point based on the inspection state and the operation data of each inspection point comprises:

for each inspection point, if the inspection state of the inspection point is not inspection, determining a preset risk word corresponding to the inspection point as a first risk word of the inspection point;

and if the inspection state of the inspection point is inspection or inspection generation, comparing the operation data of the inspection point with the statistical indexes of the inspection point to obtain a first risk word of the inspection point.

7. The method according to claim 1, wherein the determining the first estimation value and the second estimation value of the patrol point according to the risk factors of the risk words of the patrol point at each preset risk level comprises:

for each inspection point, acquiring the risk importance of the inspection point;

determining a first estimation value of the inspection point according to the risk importance of the inspection point and first risk factors of risk words of the inspection point under each preset risk level;

and determining a second estimated value of the patrol point according to the risk importance of the patrol point and second risk factors of the risk words of the patrol point under the preset risk levels.

8. The method according to claim 1, wherein before the determining the risk level of the data center using a pre-constructed risk level spatial coordinate system based on the first and second estimates of each of the patrol points, the method further comprises:

establishing a risk level space coordinate system, wherein the abscissa of the risk level space coordinate system represents the first estimation value, and the ordinate of the risk level space coordinate system represents the second estimation value;

and carrying out region division on the risk level space coordinate system according to the shape of a Chinese character 'mi', and obtaining a coordinate region corresponding to each preset risk level.

9. A risk inspection device for a data center, the device comprising:

10. A computer storage medium, wherein at least one instruction or at least one program is stored, and the at least one instruction or the at least one program is loaded and executed by a processor to implement the risk inspection method for a data center according to any one of claims 1 to 8.