CN107105028B

CN107105028B - Computer lab environment intelligent regulation system based on cloud calculates

Info

Publication number: CN107105028B
Application number: CN201710253863.3A
Authority: CN
Inventors: 董惠良; 王正敏; 杜旋; 姜学峰
Original assignee: China Tobacco Zhejiang Industrial Co Ltd
Current assignee: China Tobacco Zhejiang Industrial Co Ltd
Priority date: 2017-04-18
Filing date: 2017-04-18
Publication date: 2020-04-28
Anticipated expiration: 2037-04-18
Also published as: CN107105028A

Abstract

The invention proposes a cloud computing-based computer room environment intelligent adjustment system. Aiming at the distributed computer room system, the present invention establishes a cloud platform for intelligent adjustment of the computer room environment. The cloud platform can perform an overall judgment of the environmental state for each computer room based on the environmental index monitoring data collected from each front-end computer room, and has The environmental regulation means and facilities are issued by remote control. In this way, on the one hand, the software and hardware costs of building an integrated environmental control platform for each computer room can be saved, and the unification of environmental control standards and the upgrading of environmental control strategies can be easily realized. Algorithms implement refined control strategies, improve the intelligence of environmental regulation, and increase flexibility and adaptability.

Description

Computer lab environment intelligent regulation system based on cloud calculates

Technical Field

The invention relates to the field of communication machine room equipment, in particular to a machine room environment intelligent adjusting system based on cloud computing.

Background

The communication machine room is internally provided with various important electronic facilities such as a server, communication equipment, network management equipment, a power supply and the like, the actual working performance and the service life of the electronic facilities are closely related to the environment where the electronic facilities are located, and meanwhile, the good environment of the communication machine room is kept, so that the safety of the machine room is guaranteed, and the physical health of workers in the machine room is maintained.

Common environmental indexes of a communication machine room comprise temperature, relative humidity, temperature change rate, air cleanliness, harmful gas concentration, noise and the like. The temperature is the key that communication equipment in a machine room can run reliably for a long time, the occurrence probability of electric drift and breakdown of an integrated circuit is aggravated when the ambient temperature is too high, electronic elements such as a capacitor are damaged, the failure rate is increased, the service life is shortened, and even equipment is damaged directly, a large amount of heat can be released by equipment such as a server, communication equipment and a power supply in the machine room in the running process of the equipment, and the rise of the ambient temperature is restrained by means of cooling measures; conversely, the normal operation of the electronic equipment is also negatively affected by the excessively low ambient temperature of the machine room. The temperature change rate of the communication machine room can be stably changed within a preset allowable range; if the temperature change is too severe, the electronic component is easy to have poor contact due to expansion caused by heat and contraction caused by cold. The operation of communication equipment is influenced by overhigh and overlow relative humidity of a machine room, and equipment is easily short-circuited when the relative humidity is overhigh, poor contact is caused, and the magnetic permeability of a magnetic material is influenced; if the relative humidity is too low, static electricity is easy to generate and accumulate, which brings obstruction to the mechanical movement of the equipment and is not beneficial to the health of workers. The air cleanliness is the content of dust particles in the air; because the equipment in the machine room runs in a long-term electrified mode, dust particles are easy to adsorb on a shell, an electronic element and an integrated circuit of the equipment, circuit insulativity is reduced, and heat dissipation of the equipment is hindered; therefore, the content of dust particles in the machine room cannot exceed 1-0.75 mg/cubic meter. Some harmful substances contained in electronic components in the working process of equipment in a machine room slowly release and volatilize, so that harmful gases such as sulfur dioxide, hydrogen sulfide, nitrogen dioxide, ammonia gas, chlorine gas, hydrochloric acid, hydrofluoric acid, ozone and the like are generated, the equipment is corroded due to too high concentration of the gases, serious chronic adverse effects are generated on the bodies of workers, and even potential hazards in the safety aspect exist, so that the harmful substances are required to be kept within an allowable content. Computer unit, power, air conditioner, fan in the computer lab equipment constantly make sound under operating condition, and the noise in the computer lab should not exceed 35dB, otherwise the staff is in the environment that the noise exceeds standard for a long time and can harm auditory nerve, arouses the mood dysphoria, reduces work efficiency.

The adjusting means for maintaining the proper working environment of the communication machine room mainly comprises: various electronic facilities or cabinet boxes for placing the electronic facilities are provided with independent temperature control and cooling equipment, so that the temperature of the electronic facilities with large heating value is controlled; setting a machine room fan to adjust the temperature and the relative humidity; installing an air purifier to reduce the granularity of dust and the concentration of harmful gases; the temperature, the relative humidity, the dust granularity and the harmful gas concentration of the whole machine room are integrally regulated through an air conditioning ventilation system; and regulating and controlling the electronic equipment to enter a sleep state or reducing the working power of the equipment so as to reduce the heating and noise of the electronic equipment.

In order to maintain the machine room in a good environment state continuously and stably, and to consider the factors of saving energy consumption, improving automation degree, reducing labor management cost and the like, the machine room is expected to be capable of accurately monitoring various environmental indexes of the machine room, and then the regulation means introduced above is automatically controlled according to monitoring data, so that precise, real-time and full-automatic intelligent regulation of the machine room environment is realized.

The monitoring aiming at the environmental indexes of the machine room can be realized by utilizing the self-carried environmental parameter sensor of the electronic equipment or the machine cabinet of the machine room, or by specially deploying an environmental parameter sensor system in the space of the machine room. On the basis of obtaining the monitoring data provided by the environmental parameter sensors, in the prior art, the following specific implementation modes exist for intelligently adjusting the machine room environment.

The most basic intelligent regulation implementation manner in the prior art is to obtain monitoring data provided by an environmental parameter sensor by temperature control cooling equipment, a machine room fan, an air purifier and an air conditioning ventilation system of each electronic device or cabinet, and to determine the start/stop and the working state after start of the electronic device or cabinet according to the monitoring data. For example, when a temperature sensor arranged in a server cabinet senses that the temperature exceeds the standard, an air-cooling or water-cooling system of the cabinet automatically starts thermal circulation to cool the server in the cabinet; when the humidity sensor senses that the relative humidity of the machine room is too high, a machine room fan and/or an air conditioning ventilation system are started to dehumidify the machine room; when the air granularity sensor or the abnormal gas concentration sensor senses the abnormal index in the aspect of air quality, the air purifier and/or the air conditioner ventilation system are started to realize purification and air exchange. The realization method has the defects that the adjusting means and facilities of various machine room environments are respectively in a political view, the coordination management is lacked, the repeated operation and the over-adjustment of various facilities are difficult to avoid, and the whole machine room environment is difficult to be maintained in a stable state. For example, when an abnormal index in the air quality is sensed, the air purifier starts air purification operation after obtaining the index, and the air conditioning and ventilating system also starts ventilation and ventilation after obtaining the abnormal index, and the air conditioning and ventilating system works simultaneously to cause unnecessary repeated operation, so that energy consumption and working noise are increased, and moreover, the work of the air conditioning and ventilating system also often causes changes in the ambient temperature and relative humidity, so that unnecessary fluctuation is caused to the whole environment.

Another implementation manner in the prior art is to establish a machine room environment comprehensive control platform, where the platform obtains monitoring data of various environmental indexes collected by all environmental parameter sensors in a machine room range, comprehensively judges the current environmental state of the machine room by summarizing and analyzing all the monitoring data, and can call a corresponding environment regulation scheme according to the specific situation of environmental abnormality, and comprehensively control various regulation means such as a temperature control and cooling device, a machine room fan, an air purifier, an air conditioning and ventilation system according to the environment regulation scheme. The comprehensive control platform for the machine room environment enables the environment adjusting mode to be optimized integrally, for example, when the temperature, the relative humidity and the air granularity of the machine room are abnormal and the abnormal state exists in a larger range of the machine room space, the air conditioning ventilation system can be started preferentially for adjustment; and if only the air granularity is abnormal, the air purifier can be preferentially adopted to adjust a single environmental index so as to avoid fluctuation of temperature and humidity and achieve optimization of energy efficiency. However, the realization of the comprehensive control platform for the machine room environment requires the deployment of special console hardware and corresponding control software in the machine room, which increases the cost for the construction of the machine room; and the limited environment regulation scheme is called, so that the flexibility is lacked, and the method cannot be completely suitable for the complex and changeable environment state of the machine room.

In order to meet the application requirements of large enterprises, organizations and organizations or to realize large network communication services, a distributed computer room system, for example, a network topology structure formed by collectively networking hundreds of computer rooms distributed in various places, is often required to be built. Software and hardware required for configuring a machine room environment comprehensive control platform for each machine room respectively can bring significant cost burden, and the execution of a unified machine room environment standard is difficult to guarantee.

Disclosure of Invention

In view of the above problems in the prior art, the present invention provides an intelligent regulation system for a computer room environment based on cloud computing. The cloud platform for intelligently adjusting the machine room environment is established for a distributed machine room system, can perform overall judgment on the environment state of each machine room based on environment index monitoring data acquired from each front-end machine room, and issues remote control to environment adjusting means and facilities of the machine room. Therefore, on one hand, the software and hardware cost of the environment comprehensive control platform is built in each machine room respectively, the unification of the environment control standards and the upgrading of the environment control strategy are achieved conveniently, on the other hand, the fine control strategy can be achieved by adopting various complex algorithms through cloud computing, the intelligent degree of environment adjustment is improved, and the flexibility and the adaptability are improved.

The invention provides a machine room environment intelligent adjusting system based on cloud computing, which is characterized by comprising the following components:

the environmental parameter sensor monitoring network is arranged in each front-end machine room in the distributed machine room system; the system comprises a plurality of nodes, wherein each node is provided with a plurality of types of environmental parameter sensors for sensing environmental index monitoring data; the environmental parameter sensor establishes data links according to network organization rules of a self-organization protocol, and uploads sensed environmental index monitoring data to a machine room data transmission device serving as a root sink node through the data links after adding the environmental index monitoring data into marking fields such as acquisition time, node marks and the like;

the computer room data transmission device is provided with a wireless data interface compatible with a self-organizing protocol, so that the computer room data transmission device is accessed to the environmental parameter sensor monitoring network and is used as a root sink node of the network to obtain environmental index monitoring data sensed by all the environmental parameter sensors of the environmental parameter sensor monitoring network; the system is connected with the Internet through an Internet interface, and the environmental index monitoring data are uploaded to a cloud computing and analyzing platform through the Internet; receiving an environment regulation control instruction and other instructions and data sent by the cloud environment intelligent regulation platform;

the cloud computing and analyzing platform is used for receiving, storing and managing the environmental index monitoring data uploaded remotely by the machine room data transmission device; running an analysis algorithm of the environmental index monitoring data, analyzing the environmental index monitoring data, and determining the characteristics of the environmental state of the machine room;

the cloud environment intelligent adjusting platform is used for determining a control strategy for realizing intelligent adjustment of the machine room environment by the environment adjusting facilities in each machine room according to the characteristics of the machine room environment state determined by the cloud computing and analyzing platform, generating an environment adjusting control instruction aiming at the environment adjusting facilities based on the control strategy, and issuing the environment adjusting control instruction to the machine room data transmission device through the Internet;

the computer room environment adjusting controller is used for receiving an environment adjusting control command from the computer room data transmission device, determining an environment adjusting facility corresponding to the command, performing format matching conversion on the command, converting the command into a remote control signal capable of being executed by an internal circuit of the environment adjusting facility, and then sending the remote control signal to a computer room environment adjusting facility remote control interface;

and the machine room environment adjusting facility remote control interface is arranged on each environment adjusting facility and used for receiving the remote control signal and correspondingly controlling the work of each environment adjusting facility.

Preferably, the computer room data transmission device uploads the environmental index monitoring data in a periodic intermittent uploading mode, and each data acquisition and uploading period can be divided into a data acquisition interval and an uploading interval; in a data acquisition interval, a machine room data transmission device is used as a root sink node to continuously receive environment index monitoring data uploaded by an environment parameter sensor monitoring network, cache the received environment index monitoring data, analyze whether the acquisition time of the received environment index monitoring data belongs to the data acquisition interval or not, and add an interval identifier representing the data acquisition interval to the environment index monitoring data if the acquisition time of the received environment index monitoring data belongs to the data acquisition interval; and in an uploading interval, the machine room data transmission device uploads the environmental index monitoring data with the interval identification of the data acquisition interval.

Preferably, the cloud computing analysis platform includes:

the environmental index monitoring database module is used for receiving environmental index monitoring data uploaded by the machine room data transmission device in each uploading interval and storing the environmental index monitoring data in the environmental index monitoring database; the environment index monitoring database module establishes an independent environment index monitoring data form for each front-end computer room in the distributed computer room system, and the environment index monitoring data of each node is stored, called and managed by taking the environment index type, the interval identification of the data acquisition interval and the node mark as indexes in the form;

the environment pattern generation module is used for calling the environment index monitoring data of all the nodes on each data acquisition interval from the environment index monitoring database module so as to generate an environment index distribution pattern for each data acquisition interval;

the environment pattern caching module is used for caching the environment index distribution pattern generated for each data acquisition interval;

the environment pattern classifying module is used for acquiring an environment index distribution pattern of a current data acquisition interval and environment index distribution patterns of a plurality of data acquisition intervals which are previous in time from the environment pattern caching module, classifying the environment index distribution pattern corresponding to the current data acquisition interval according to the change of the environment index distribution pattern among the data acquisition intervals, and the classified types comprise a stable environment pattern, a local change environment pattern and a global change environment pattern;

and the environment pattern feature extraction module is used for classifying the pattern according to the environment pattern classification module aiming at the environment index distribution pattern of the current data acquisition interval, and extracting the symbolic feature value in the pattern in a mode corresponding to the classification to be used as the feature of the machine room environment state.

Further preferably, the environment pattern generation module fills the empty pixel points of the pattern with numerical values by adopting a spatial interpolation mode, a temporal interpolation mode or a combination mode of the spatial interpolation mode and the temporal interpolation mode for the empty pixel points generated on the corresponding environment index distribution pattern due to the lack of the environment index monitoring data of the node on the data acquisition interval.

Further preferably, the environmental pattern generation module obtains environmental index distribution patterns of one or more preceding and following data acquisition intervals adjacent to the current data acquisition interval in terms of time, and calculates an average pixel value in terms of time for each pixel point according to pixel values of each pixel point in the environmental index distribution patterns adjacent in terms of time; then, obtaining a pixel point which is adjacent to the empty pixel point on the environment index distribution pattern of the current data acquisition interval and has an environment index monitoring data value, and calculating the pixel point change ratio of the pixel point adjacent to the empty pixel point relative to the average pixel value; and according to the average value of the pixel value change ratios, calculating the filling value of the empty pixel point by multiplying the average pixel value of the pixel points positioned at the empty pixel point position in the temporally adjacent environment index distribution pattern by the average value of the pixel value change ratios.

Preferably, the environment pattern classifying module performs pixel difference operation on the environment index distribution pattern of the current data acquisition interval and the environment index distribution pattern of the preceding and/or following data acquisition interval adjacent in time respectively to obtain a difference absolute value on each pixel point; calculating the average value of the absolute values of the difference values on each pixel point according to the obtained absolute values of the differences, judging whether the number of the pixel points of which the average value of the absolute values of the difference values is larger than a preset threshold value is smaller than a stable threshold value or not, and if the average value of the absolute values of the difference values is smaller than the stable threshold value, classifying the environmental index distribution pattern of the current data acquisition interval into a stable environmental pattern; if the number of the pixel points of which the average value of the absolute value of the difference value is greater than the preset threshold value is not less than the stable threshold value, judging whether the number of the pixel points of which the average value of the absolute value of the difference value is greater than the preset threshold value is greater than a global change threshold value or not, and if so, classifying the environmental index distribution pattern of the current data acquisition interval into a global change environmental pattern; if the number of the pixel points of which the average value of the absolute value of the difference value is greater than the preset threshold is greater than the stable threshold and is not greater than the global change threshold, further analyzing the block concentration degree of the pixel points of which the average value of the absolute value of the difference value is greater than the preset threshold, and determining whether the environmental index distribution pattern of the current data acquisition interval is included in the global change environment pattern or the local change environment pattern according to the block concentration degree.

More preferably, when the number of the pixels of which the average value of the absolute values of the difference values is greater than the predetermined threshold is greater than the stability threshold and is not greater than the global change threshold, the environment pattern classification module establishes a blank template in which the pixel points and the environment index distribution pattern are in one-to-one correspondence; if yes, marking a blank template pixel point corresponding to the current pixel point position as 1; if not, marking the blank template pixel point corresponding to the current pixel point position as 0; for the pixel points of which the average value of the absolute value of the difference value is not more than the preset threshold value, marking the corresponding blank template pixel points as 0; after each pixel point of the blank template is marked in sequence, the blank template is converted into a binary pattern; counting the number of pixel points with the median value of 1 in the binary pattern, and judging whether the number of the pixel points is greater than a region change judgment threshold value or not; if the change is larger than the area change judgment threshold, the environmental index distribution pattern of the current data acquisition interval is classified into a local change environmental pattern, and if the change is not larger than the area change judgment threshold, the environmental index distribution pattern of the current data acquisition interval is classified into a global change environmental pattern.

Preferably, the environmental pattern feature extraction module divides the pattern into a plurality of blocks according to the condition that the environmental index distribution pattern of the current data acquisition interval is classified as a stable environmental pattern, calculates the average value of the values of all the pixel points in each block, and extracts the average value of each block as the symbolic feature value of the pattern;

the environmental pattern feature extraction module divides the environmental index distribution pattern of the current data acquisition interval into a plurality of blocks according to the condition that the pattern is classified as a global change environmental pattern; aiming at each block, obtaining the average value of the absolute value of the difference value of each pixel point, determining the pixel points of which the average value of the absolute value of the difference value is greater than the preset threshold value in the block, and calculating the average value of the pixel values of the pixel points; extracting the average value corresponding to each block as a symbolic characteristic value of the pattern;

the method comprises the following steps that an environment pattern feature extraction module firstly obtains a binarization pattern corresponding to an environment index distribution pattern under the condition that the environment index distribution pattern of a current data acquisition interval is classified into a local change environment pattern; then, selecting a pixel point corresponding to the position marked as 1 in the binary pattern from the environment index distribution pattern; all the pixel points with adjacent relation among the pixel points are classified into a block as a local change block; calculating an average value of pixel values of pixel points in the local change block; extracting the average value corresponding to each local variation block as a symbolic characteristic value of the pattern; and, divide the pattern into several blocks evenly; aiming at each uniformly divided block, obtaining pixel points corresponding to the positions marked as 0 in the binary pattern; and calculating the average value of the pixel values of the pixel points corresponding to the positions marked as 0 in the binary pattern in each uniformly divided block, thereby extracting the average value corresponding to each uniformly divided block and also taking the average value as the symbolic characteristic value of the pattern.

Preferably, the cloud environment intelligent regulation platform includes:

the environment state abnormity judgment module is used for acquiring a symbolic characteristic value of each type of environment index distribution pattern and judging whether the symbolic characteristic value is within a preset normal environment threshold value range; determining a block corresponding to the symbolic feature value which is not in the normal environment threshold range in the pattern, and taking the block as an abnormal block;

the environment abnormal pattern recognition module is used for recognizing the machine room environment abnormal pattern according to the classification type of the environment index distribution pattern of each type, the symbolic characteristic value judged to be abnormal and the abnormal block corresponding to the symbolic characteristic value;

the regulating facility parameter registration module is used for registering the type, the number, the installation position and the adjustable working parameters of the environment regulating facilities of each machine room;

the regulation and control strategy determining module is used for determining an environment regulation and control strategy aiming at each environment regulation facility of the machine room according to the abnormal environment mode of the machine room identified by the abnormal environment mode identifying module and the registration of the regulation and control facility parameter registering module;

the regulation and control instruction generation module is used for generating an environment regulation and control instruction aiming at each environment regulation facility of the machine room according to the environment regulation and control strategy determined by the regulation and control strategy determination module; and sending the environmental regulation control command to the machine room data transmission device of the front-end machine room through the network.

The improvement and the beneficial effects of the invention are that a cloud computing platform for machine room environment state analysis and intelligent regulation strategy is established, and aiming at each front-end machine room in a distributed machine room system, the unified remote cloud platform can be used for realizing the control of environment regulation facilities in the machine room, so that a set of special intelligent regulation system is not required to be established for each machine room, and the software and hardware cost is greatly saved; moreover, the cloud computing platform has sufficient computing resources and complex algorithm support, so that fine environmental state analysis can be realized, a targeted personalized regulation and control strategy is adopted for the unique environmental state and the change trend of each machine room in each time period, and the problems of low flexibility and poor applicability caused by the realization of control based on a fixed plan are solved; the invention takes pattern analysis as a basic means for analyzing the environmental state of the machine room, is beneficial to shielding the difference of each machine room in facility configuration and spatial layout, realizes a unified and effective analysis method and improves the compatibility of an analysis algorithm.

Drawings

FIG. 1 is a schematic diagram of the overall structure of an intelligent regulation system for a machine room environment according to a preferred embodiment of the present invention;

FIG. 2 is a schematic diagram of an environmental parameter sensor monitoring network sensor layout according to a preferred embodiment of the present invention;

fig. 3 is a schematic diagram of an uplink transmission working cycle of the machine room data transmission device according to the preferred embodiment of the present invention;

FIG. 4 is a schematic diagram of a cloud computing analysis platform according to a preferred embodiment of the present invention;

FIG. 5 is a schematic view of an environmental indicator distribution pattern pixel in accordance with a preferred embodiment of the present invention;

FIGS. 6A and 6C are schematic diagrams of the distribution pattern of the environmental indicators according to the preferred embodiment of the present invention;

fig. 6B and 6D are schematic diagrams of binarization patterns of the environmental index distribution pattern according to the preferred embodiment of the present invention;

FIG. 7A is a block diagram illustrating the stable environment pattern feature value extraction according to the preferred embodiment of the present invention;

FIG. 7B is a block diagram illustrating the partitioning of the global change environment pattern feature value extraction according to the preferred embodiment of the present invention;

FIG. 7C is a schematic diagram illustrating a partition of a feature value extraction block of a pattern in a local variation environment according to an embodiment of the present invention;

fig. 8 is a schematic structural diagram of a cloud environment intelligent adjustment platform according to a preferred embodiment of the present invention.

Detailed Description

The technical scheme of the invention is further specifically described by the following embodiments.

The invention provides a machine room environment intelligent adjusting system based on cloud computing. The cloud platform judges the whole of the execution environment state of each machine room according to the environment index monitoring data collected from each front-end machine room of the distributed machine room system, and issues remote control to the environment adjusting means and facilities of the machine room. The cloud platform executes cloud computing for analyzing the environmental state of the machine room on the basis of the environmental index monitoring data, and then a control strategy for intelligently adjusting the environment of the machine room is realized.

For a distributed computer room system, the internal environmental factors of each computer room are very complex, for example, the types, powers, heating values and quantities of electronic facilities such as servers and the like in different computer rooms are different, and the spatial arrangement and the working state of various electronic facilities are different. These factors have direct influence on the temperature, relative humidity, ventilation effect, dust particle distribution and purification capacity and noise in the machine room, and thus, there is diversity in the specific factors to be considered in the analysis of the machine room environment. The cloud platform is to realize accurate analysis of the environmental state of the machine rooms according to the environmental index monitoring data, and cannot ignore essential differences of the machine rooms on the environmental influence factors. However, for the remote cloud platform, it is difficult to obtain or measure the specific conditions of the environmental influence factors inside each machine room, and the facility configuration and the spatial arrangement of the machine room are often adjusted manually. Therefore, the cloud platform establishes the environment index distribution pattern for each machine room on the basis of the environment index monitoring data, and realizes the individual analysis of the unique environment state of each machine room according to the distribution characteristics of the environment index distribution pattern and the change of the environment index distribution pattern along with time.

Referring to fig. 1, the intelligent regulation system for a computer room environment based on cloud computing specifically includes: the system comprises an environmental parameter sensor monitoring network 101, a machine room data transmission device 102, a cloud computing analysis platform 103, a cloud environment intelligent adjusting platform 104, a machine room environment adjusting controller 105 and a machine room environment adjusting facility remote control interface 106.

The environmental parameter sensor monitoring network 101 is arranged in each front-end machine room I-III in the distributed machine room system, and each environmental parameter sensor in each front-end machine room I-III is added to the environmental parameter sensor monitoring network 101 in the front-end machine room. The environmental parameter sensor monitoring network 101 includes various types of environmental parameter sensors for sensing environmental index monitoring data such as temperature, relative humidity, air dust particle content, harmful gas concentration, noise decibel, etc., such as a temperature sensor, a relative humidity sensor, an air dust particle detection sensor, a harmful gas concentration sensor, a noise decibel sensor. As a preferable scheme, each environmental parameter sensor adopts a wireless sensor for realizing a wireless data transmission function based on a self-organizing protocol such as ZigBee, and the environmental parameter sensors mutually establish data links according to network organization rules of the self-organizing protocol such as ZigBee; the environmental parameter sensor as the bottom node adds the sensed environmental index monitoring data into the marking fields such as acquisition time, node marks and the like, and uploads the environmental index monitoring data to the environmental parameter sensor as the sink node; under a multilayer network architecture, the sink nodes upload the environmental index monitoring data to an environmental parameter sensor serving as a sink node of an upper layer, so that the environmental parameter sensor monitoring network 101 is formed; moreover, the machine room data transmission device 102 is used as a root sink node of the environmental parameter sensor monitoring network 101 formed based on the self-organizing protocol such as ZigBee, so that the machine room data transmission device 102 can finally obtain the environmental index monitoring data sensed by all the environmental parameter sensors of the whole environmental parameter sensor monitoring network 101. The machine room data transmission device 102 may further calculate a temperature change rate at each node according to the temperature sensed by the temperature sensor in each sampling period, and use the temperature change rate as the sensed environmental index monitoring data. In terms of the spatial layout of the environmental parameter sensors, it is preferable that the environmental parameter sensors are evenly distributed in the entire room space. For example, as shown in fig. 2, in the machine room space region Z, each black dot represents an environmental parameter sensor cluster, and each environmental parameter sensor cluster is composed of a temperature sensor, a relative humidity sensor, an air dust and particle detection sensor, a harmful gas concentration sensor, and a noise decibel sensor; therefore, the environmental parameter sensor clusters are uniformly distributed in the space range of the machine room space region Z, and the environmental parameter sensor clusters are required to have enough distribution density, so that the environmental index monitoring data provided by the environmental parameter sensor clusters have enough sampling rate relative to the machine room space, and the environmental state distribution of the machine room can be comprehensively reflected.

The machine room data transmission device 102 is a communication interface component for remotely connecting the machine room with the cloud platform through the internet and realizing uplink and downlink data transmission. As described above, on one hand, the machine room data transmission device 102 has a wireless data interface compatible with self-organizing protocols such as ZigBee, so that the environmental parameter sensor monitoring network 101 can be accessed and environmental index monitoring data sensed by all the environmental parameter sensors of the environmental parameter sensor monitoring network 101 can be obtained. On the other hand, the machine room data transmission device 102 is connected with the internet through an internet interface, so that bidirectional interactive data communication is realized with the cloud platform of the invention through network address configuration; the machine room data transmission device 102 may upload the environmental index monitoring data to the cloud platform, and receive an environmental regulation control instruction and other instructions and data sent by the cloud platform.

The machine room data transmission device 102 adopts a periodic intermittent uploading mode for uploading the environmental index monitoring data. As shown in fig. 3, a data collection and upload cycle of the data transmission device 102 in the computer room may be divided into a data collection interval T11 and an upload interval T21. In a data acquisition interval T11, the machine room data transmission device 102 serves as a root sink node to continuously receive environmental index monitoring data uploaded by other nodes of the environmental parameter sensor monitoring network 101; the machine room data transmission device 102 buffers the received environmental index monitoring data, and analyzes whether the acquisition time of the received environmental index monitoring data belongs to the data acquisition interval T11, if so, adds an interval identifier indicating the data acquisition interval T11 to the environmental index monitoring data; furthermore, in the upload interval T21, the machine room data transmission device 102 uploads the environmental index monitoring data with the interval identifier of the data acquisition interval T11 to the cloud-end platform. For the environmental index monitoring data received by the machine room data transmission device 102 in the uploading interval T21, the machine room data transmission device 102 issues a delay instruction to the nodes providing the environmental index monitoring data, and requests the nodes to delay the sensing and uploading of the environmental index monitoring data again in the next period after the uploading interval T21 is finished. After the uploading interval T21 is ended, the machine room data transmission device 102 enters the data acquisition interval T12 and the uploading interval T22 of the next period, and so on, successively enters the data acquisition interval T13 and the uploading interval T23, and so on.

The cloud platform of the invention is described below in detail, specifically, the intelligent machine room environment adjusting system based on cloud computing establishes a server at the cloud, and the server at the cloud can run a relatively complex intelligent analysis algorithm and a remote control system, so that management, analysis and storage of environment index monitoring data are realized, and an intelligent machine room environment adjusting function is provided for each front-end machine room. The cloud platform of the invention specifically comprises a cloud computing analysis platform 103 and a cloud environment intelligent regulation platform 104, wherein the two platforms can be realized by physically different servers or server groups, and can also be respectively operated on the same server or server group to realize mutually independent virtual platforms.

The cloud computing and analyzing platform 103 is used for receiving, storing and managing the environmental index monitoring data uploaded remotely by the machine room data transmission device 102; and on the basis of the environmental index monitoring data, operating an analysis algorithm of the environmental index monitoring data to determine the environmental state of the machine room. As shown in fig. 4, the cloud computing analysis platform 103 specifically includes: the environment index monitoring system comprises an environment index monitoring database module 103A, an environment pattern generating module 103B, an environment pattern caching module 103C, an environment pattern classifying module 103D and an environment pattern feature extracting module 103E.

The environmental index monitoring database module 103A is configured to receive environmental index monitoring data uploaded by the machine room data transmission device 102 in each uploading interval T21, T22, T23, and store the environmental index monitoring data in the environmental index monitoring database. Other modules of the cloud computing analysis platform 103 can call data in the environment index monitoring database at any time according to the needs of the analysis algorithm. The environment index monitoring database module 103A establishes an independent environment index monitoring data form for each front-end computer room in the distributed computer room system; in the database form, the environmental index type is used as a primary index, and environmental index monitoring data about temperature, relative humidity, air dust particle content, harmful gas concentration and noise decibel can be respectively called; taking respective section identifications of the data acquisition sections T11, T12, T13 and the like as secondary indexes, so that environmental index monitoring data respectively possessed by the front-end computer room in the data acquisition sections T11, T12 and T13 … … can be called out; further, with the node labels as three-level indexes, the environmental index monitoring data of each node of the environmental parameter sensor monitoring network 101 on each data collection interval T11, T12, T13 … … can be called.

For any type of environmental index monitoring data, the environmental pattern generation module 103B retrieves the environmental index monitoring data of all nodes on each data collection interval T11, T12, T13 … … from the environmental index monitoring database module 103A, so as to generate an environmental index distribution pattern for each data collection interval T11, T12, T13 … …, as shown in fig. 5, each pixel point (black dot in fig. 5) of the pattern represents a node of the environmental parameter sensor monitoring network 101, and the value of the pixel point on the pattern is the value of the environmental index monitoring data sensed by the node in the data collection interval.

For reasons of sensing and transmission delay, data packet loss, and the like, for a certain data acquisition interval, the environmental index monitoring database module 103A may lack environmental index monitoring data of a part of nodes. For example, the pixel points marked by the dashed circles in fig. 5 lack the environmental index monitoring data of the corresponding node in the data collection interval in the environmental index monitoring database module 103A, so that a null pixel point is generated on the corresponding environmental index distribution pattern. For the empty pixel, the environment pattern generating module 103B may perform processing on the empty pixel of the pattern by adopting a spatial interpolation, a temporal interpolation, or a combination of the two, and fill the pixel with a value. Specifically, when the spatial interpolation mode is adopted, the environment pattern generation module 103B calls a pixel point adjacent to the empty pixel point and having an environment index monitoring data value, and calculates an average value of pixel values of the called adjacent pixel points as a filling value of the empty pixel point; the pixel points surrounded by the dotted line box in fig. 5 are the pixel points adjacent to the empty pixel point. When the time interpolation mode is adopted, the environment pattern generation module 103B obtains the environment index distribution patterns of one or more preceding and following data acquisition sections that are temporally adjacent to the current data acquisition section, and then calls the pixels that have the same position as the empty pixel and have the environment index monitoring data value on the adjacent environment index distribution patterns, and calculates the average value of the pixel values of the called pixels as the filling value of the empty pixel. When an interpolation mode combining space and time is adopted, the environment pattern generation module 103B obtains the environment index distribution patterns of one or more preceding and following data acquisition intervals temporally adjacent to the current data acquisition interval, and calculates the average pixel value of each pixel point temporally according to the pixel value of each pixel point in the temporally adjacent environment index distribution patterns; then, obtaining a pixel point which is adjacent to the empty pixel point on the environment index distribution pattern of the current data acquisition interval and has an environment index monitoring data value, and calculating the pixel point change ratio of the pixel point adjacent to the empty pixel point relative to the average pixel value; and according to the average value of the pixel value change ratios, calculating the filling value of the empty pixel point by multiplying the average pixel value of the pixel points positioned at the empty pixel point position in the temporally adjacent environment index distribution pattern by the average value of the pixel value change ratios.

The environmental pattern buffer module 103C is configured to buffer the environmental indicator distribution patterns generated for each data acquisition interval T11, T12, and T13 … …. Specifically, the environmental pattern buffer module 103C establishes a buffer queue for each type of environmental index monitoring data, including temperature, relative humidity, air dust particle content, harmful gas concentration, noise decibel, and the like, and sequentially buffers the environmental index distribution patterns generated by each data acquisition interval T11, T12, and T13 … … in each type of buffer queue according to the time sequence of the data acquisition interval.

The environmental pattern classifying module 103D obtains the environmental index distribution pattern of the current data acquisition interval, for example, the data acquisition interval T13, and the environmental index distribution patterns of the several data acquisition intervals preceding in time, for example, the data acquisition intervals T12 and T11, from each type of buffer queue of the environmental pattern buffer module 103C. Furthermore, the environmental pattern classification module 103D classifies the environmental index distribution pattern corresponding to the current data acquisition interval according to the change of the environmental index distribution pattern among the data acquisition intervals, and the classification types include a stable environmental pattern, a local change environmental pattern, and a global change environmental pattern.

Specifically, the pixel difference operation is performed on the environmental index distribution pattern of the current data acquisition interval T13 and the environmental index distribution patterns of the preceding and following data acquisition intervals T12 and T11 which are adjacent in time, that is, the difference calculation is performed on each pixel point on the environmental index distribution pattern of T13 and the pixel point at the same position on the environmental index distribution pattern of T12, so as to obtain the difference absolute value of each pixel point; similarly, the difference absolute value between each pixel point on the environmental index distribution pattern of T13 and the pixel point at the same position on the environmental index distribution pattern of T11 is obtained. And calculating the average value of the absolute values of the difference values on each pixel point according to the obtained difference absolute values. And judging whether the number of the pixel points of which the average value of the absolute values of the difference values is greater than a preset threshold value is less than a stable threshold value or not, and if so, classifying the environmental index distribution pattern of the current data acquisition interval T13 into a stable environmental pattern. On the contrary, if the number of the pixels of which the average value of the absolute values of the difference values is greater than the predetermined threshold is not less than the stable threshold, it is further determined whether the number of the pixels of which the average value of the absolute values of the difference values is greater than the predetermined threshold is greater than the global change threshold, and if so, the environmental index distribution pattern of the current data acquisition interval T13 is classified into the global change environmental pattern. If the number of the pixel points of which the average value of the absolute values of the differential values is greater than the preset threshold is greater than the stable threshold and is not greater than the global change threshold, further analyzing the concentration degree of the pixel points of which the average value of the absolute values of the differential values is greater than the preset threshold; recording the pixel points of which the average value of the absolute value of each difference value is greater than the preset threshold as the current pixel point, and judging whether the pixel points of which the average value of the absolute value of the difference value is greater than the preset threshold exist in the adjacent pixel points of the current pixel point; if yes, marking a blank template pixel point corresponding to the current pixel point position as 1; if not, marking the blank template pixel point corresponding to the current pixel point position as 0; for the pixel points of which the average value of the absolute value of the difference value is not more than the preset threshold value, marking the corresponding blank template pixel points as 0; after each pixel point of the blank template is marked in sequence, the blank template is converted into a binary pattern; for example, for the environmental index distribution pattern shown in fig. 6A, where the pixel points marked as black are the pixel points where the average value of the absolute values of the difference values is greater than the predetermined threshold, the obtained binarization pattern is as shown in fig. 6B according to the above method; for the environmental index distribution pattern shown in fig. 6C, the pixel points marked as black are the pixel points in which the average value of the absolute values of the difference values is greater than the predetermined threshold, and the obtained binarization pattern is shown in fig. 6D according to the above method; it can be seen that, for the pixels whose average value of the absolute value of the original difference value is greater than the predetermined threshold, only the pixels concentrated in one block are marked as 1 in the binarization pattern, so that although the total number of black pixels in fig. 6A and 6C is the same, the number of pixels marked as 1 in the binarization pattern of fig. 6B is significantly greater than that in fig. 6D; further, counting the number of pixel points with the median value of 1 in the binary pattern, and judging whether the number of the pixel points is greater than a region change judgment threshold value or not; if the environmental index distribution pattern of the current data acquisition interval T13 is greater than the area change determination threshold, the environmental index distribution pattern of the current data acquisition interval T13 is classified as a local change environmental pattern, and if the environmental index distribution pattern of the current data acquisition interval T13 is not greater than the area change determination threshold, the global change environmental pattern.

The environmental pattern feature extraction module 103E extracts the landmark feature values in the pattern in different ways according to the classification of the pattern by the environmental pattern classification module 103D for the environmental index distribution pattern of the current data acquisition interval, for example, the data acquisition interval T13. Specifically, for the case that the environmental index distribution pattern of the data acquisition interval T13 is classified as a stable environmental pattern, the pixel values of the environmental index distribution pattern are uniformly extracted as the landmark characteristic values; for example, fig. 7A shows a stable environment pattern, the pattern is divided into a plurality of blocks according to the dotted lines in the pattern, the average value of the values of the pixels in each block is calculated, and the average value of each block is extracted as the landmark feature value of the pattern. For the case that the distribution pattern of the environmental indicators of the data acquisition interval T13 is classified as a global change environmental pattern, as shown in fig. 7B, the pattern is divided into a plurality of blocks according to the dotted lines in the figure; for each block, obtaining an average value of the absolute values of the difference values of each pixel point, and determining the pixel points in the block, in which the average value of the absolute values of the difference values is greater than the predetermined threshold, for example, the black pixel points in fig. 7B are the pixel points in which the average value of the absolute values of the difference values is greater than the predetermined threshold; calculating the average value of the pixel values of the pixel points aiming at the pixel points of which the average value of the absolute value of the difference value in each block is greater than the preset threshold value; and extracting the average value corresponding to each block as a symbolic characteristic value of the pattern. In the case where the environmental index distribution pattern of the data acquisition interval T13 is classified as a locally changing environmental pattern, first, the binarizing pattern corresponding to the environmental index distribution pattern is obtained; then, selecting a pixel corresponding to the position marked as 1 in the binarization pattern from the environment index distribution pattern, for example, a black pixel in fig. 7C is a pixel corresponding to the position marked as 1 in the binarization pattern; all the pixels having an adjacent relationship among the pixels are classified into a block as a local change block, for example, as shown by a bold solid line in fig. 7C, the local change block is divided; calculating an average value of pixel values of pixel points in the local change block; extracting the average value corresponding to each local variation block as a symbolic characteristic value of the pattern; and, as shown by the dotted lines in fig. 7C, the pattern is divided into several blocks; for each block, obtaining a pixel point corresponding to a position marked as 0 in the binarization pattern, as shown by a white pixel point in fig. 7C; and calculating the average value of the pixel values of the pixel points corresponding to the positions marked as 0 in the binarization pattern in each block, thereby extracting the average value corresponding to each block and also taking the average value as the symbolic characteristic value of the pattern.

Thus, through the above processing, the cloud computing and analyzing platform 103 can generate a set of symbolic characteristic values for the environmental index distribution pattern of each type of environmental index monitoring data in the current data acquisition interval.

For each front-end computer room, the cloud environment intelligent adjusting platform 104 receives classification types and symbolic characteristic values of various types of environment index distribution patterns in the current data acquisition interval from the cloud computing and analyzing platform 103, and determines a control strategy for realizing intelligent adjustment of the computer room environment by using environment adjusting facilities such as temperature control and cooling equipment, computer room fans, air purifiers, air conditioning and ventilation systems and the like in each computer room according to the classification types and the symbolic characteristic values. As shown in fig. 8, the cloud environment intelligent adjustment platform 104 specifically includes: an environmental state anomaly judgment module 104A, an environmental anomaly pattern recognition module 104B, a regulation and control facility parameter registration module 104C, a regulation and control strategy determination module 104D, and a regulation and control instruction generation module 104E.

The environmental status anomaly determination module 104A obtains a symbolic feature value of each type of environmental index distribution pattern, and determines whether the symbolic feature value is within a predetermined normal environmental threshold range; and determining a block corresponding to the symbolic feature value which is not in the normal environment threshold range in the pattern, and taking the block as an abnormal block.

The environmental abnormal pattern recognition module 104B recognizes the environmental abnormal pattern of the machine room according to the classification type of the environmental index distribution pattern of each type, the symbolic characteristic value determined to be abnormal, and the abnormal block corresponding to the symbolic characteristic value. Specifically, the environmental abnormal pattern recognition module 104B first determines the number of corresponding abnormal blocks according to all the abnormal landmark characteristic values, and recognizes the machine room environmental abnormal pattern as a global abnormality when the number of the abnormal blocks is greater than or equal to a global abnormal threshold; conversely, if the number of abnormal blocks is less than the global abnormal threshold, the abnormal pattern of the machine room environment is identified as a local abnormality. And if the environmental index distribution pattern is classified as a local change environmental pattern, the environmental abnormal pattern recognition module 104B further determines whether the abnormal block therein includes a local change block, and if so, further recognizes the environmental abnormal pattern of the machine room as a local change abnormality. Under the condition that the environmental index distribution pattern is classified as a global change environmental pattern, if the machine room environmental abnormal pattern is identified as global abnormal in the previous step, the machine room environmental abnormal pattern is further identified as global change abnormal; and if the machine room environment abnormal mode is identified as local abnormal in the previous step, further identifying the machine room environment abnormal mode as global change local abnormal.

The control facility parameter registration module 104C registers the type, number, installation location, and adjustable operating parameters of the environment control facilities in each machine room. For the facilities with locally effective adjusting functions such as temperature control and cooling equipment, machine room fans and the like, the adjustable working parameters comprise the parameters of the temperature control and cooling equipment, the power of the machine room fans and the effective space range of the machine room fans; for globally effective environmental conditioning facilities such as air purifiers, air conditioning ventilation systems, etc., the adjustable operating parameters include power parameters thereof.

The regulation and control policy determination module 104D determines the environment regulation and control policy for each environment regulation facility of the machine room according to the abnormal environment pattern of the machine room identified by the abnormal environment pattern identification module 104B and the registration of the regulation and control facility parameter registration module 104C. For example, for a global exception, a globally valid environment adjustment facility is invoked with priority; for local exception, the local effective environment regulation facility is preferably called; in the mode identified as global change local anomaly, the local effective environment regulation facility is called, and the global effective environment regulation facility is called at the same time, so that the local anomaly is prevented from diffusing along with global change; for the mode of local abnormal change, the adjusting power of the local effective environment adjusting facility is increased relative to the local abnormal so as to inhibit the local abnormal change trend; and for global change abnormity, the adjusting power of the globally effective environment adjusting facility is increased relative to the global abnormity so as to inhibit the global abnormal change trend.

The regulation and control instruction generation module 104E generates an environment regulation control instruction for each environment regulation facility of the machine room according to the environment regulation and control policy determined by the regulation and control policy determination module 104D. And the module issues the instruction to the machine room data transmission device 102 of the front-end machine room through the network.

The machine room environment adjustment controller 105 located in the front end machine room receives the environment adjustment control command from the machine room data transmission device 102, determines the environment adjustment facility to which the command is directed, performs format matching conversion on the command, converts the command into a remote control signal that can be executed by an internal circuit of the environment adjustment facility, and then sends the remote control signal to the machine room environment adjustment facility remote control interface 106. The machine room environment adjusting facility remote control interface 106 is installed on each temperature control cooling device, a machine room fan, an air purifier, an air conditioning ventilation system and other environment adjusting facilities, and is used for receiving remote control signals and correspondingly controlling the opening and closing of each environment adjusting facility and working parameters such as working power and working gear of each environment adjusting facility.

The above embodiments are only for illustrating the invention and are not to be construed as limiting the invention, and those skilled in the art can make various changes and modifications without departing from the spirit and scope of the invention, therefore, all equivalent technical solutions also belong to the scope of the invention, and the scope of the invention is defined by the claims.

Claims

1. The utility model provides a computer lab environment intelligent regulation system based on cloud, its characterized in that, this system includes:

the environmental parameter sensor monitoring network is arranged in each front-end machine room in the distributed machine room system; the system comprises a plurality of nodes, wherein each node is provided with a plurality of types of environmental parameter sensors for sensing environmental index monitoring data; the environmental parameter sensor establishes data links according to network organization rules of a self-organization protocol, and uploads sensed environmental index monitoring data to a machine room data transmission device serving as a root sink node through the data links after adding acquisition time and a node mark marking field;

the remote control interfaces of the machine room environment adjusting facilities are arranged on the environment adjusting facilities and used for receiving remote control signals and correspondingly controlling the work of the environment adjusting facilities;

the computer room data transmission device uploads the environmental index monitoring data in a periodic intermittent uploading mode, and each data acquisition and uploading period can be divided into a data acquisition interval and an uploading interval; in a data acquisition interval, a machine room data transmission device is used as a root sink node to continuously receive environment index monitoring data uploaded by an environment parameter sensor monitoring network, cache the received environment index monitoring data, analyze whether the acquisition time of the received environment index monitoring data belongs to the data acquisition interval or not, and add an interval identifier representing the data acquisition interval to the environment index monitoring data if the acquisition time of the received environment index monitoring data belongs to the data acquisition interval; in an uploading interval, the machine room data transmission device uploads the environmental index monitoring data with the interval identification of the data acquisition interval;

the cloud computing analysis platform comprises:

the environment pattern classifying module is used for acquiring an environment index distribution pattern of a current data acquisition interval and environment index distribution patterns of a plurality of data acquisition intervals which are previous in time from the environment pattern caching module, and classifying the environment index distribution pattern corresponding to the current data acquisition interval according to the change of the environment index distribution pattern among the data acquisition intervals, wherein the classified types comprise a stable environment pattern, a local change environment pattern or a global change environment pattern;

the environment pattern feature extraction module is used for classifying the pattern according to the environment pattern classification module aiming at the environment index distribution pattern of the current data acquisition interval, and extracting a symbolic feature value in the pattern in a mode corresponding to the classification to be used as the feature of the machine room environment state;

the environment pattern generation module fills numerical values for the empty pixel points of the pattern by adopting a space interpolation mode, a time interpolation mode or a combination mode aiming at the empty pixel points generated on the corresponding environment index distribution pattern due to the lack of the environment index monitoring data of the nodes on the data acquisition interval;

when an interpolation mode combining space and time is adopted, the environment pattern generation module obtains environment index distribution patterns of one or more previous and next data acquisition intervals adjacent to the current data acquisition interval in terms of time, and calculates the average pixel value in terms of time for each pixel point according to the pixel value of each pixel point in the environment index distribution patterns adjacent in terms of time; then, obtaining a pixel point which is adjacent to the empty pixel point on the environment index distribution pattern of the current data acquisition interval and has an environment index monitoring data value, and calculating the pixel point change ratio of the pixel point adjacent to the empty pixel point relative to the average pixel value; calculating the filling value of the empty pixel point according to the average value of the pixel point change ratios and the average pixel value of the pixel point positioned at the empty pixel point position in the temporally adjacent environment index distribution pattern multiplied by the average value of the pixel point change ratios;

the environment pattern classifying module carries out pixel difference operation on the environment index distribution pattern of the current data acquisition interval and the environment index distribution pattern of the previous and/or subsequent data acquisition intervals which are adjacent in time respectively to obtain a difference absolute value on each pixel point; calculating the average value of the absolute values of the difference values on each pixel point according to the obtained absolute values of the differences, judging whether the number of the pixel points of which the average value of the absolute values of the difference values is larger than a preset threshold value is smaller than a stable threshold value or not, and if the average value of the absolute values of the difference values is smaller than the stable threshold value, classifying the environmental index distribution pattern of the current data acquisition interval into a stable environmental pattern; if the number of the pixel points of which the average value of the absolute value of the difference value is greater than the preset threshold value is not less than the stable threshold value, judging whether the number of the pixel points of which the average value of the absolute value of the difference value is greater than the preset threshold value is greater than a global change threshold value or not, and if so, classifying the environmental index distribution pattern of the current data acquisition interval into a global change environmental pattern; if the number of the pixel points of which the average value of the absolute values of the difference values is greater than the preset threshold is greater than the stable threshold and is not greater than the global change threshold, further analyzing the block concentration degree of the pixel points of which the average value of the absolute values of the difference values is greater than the preset threshold, and determining whether the environmental index distribution pattern of the current data acquisition interval is included in a global change environment pattern or a local change environment pattern according to the block concentration degree;

when the number of the pixel points of which the average value of the absolute values of the difference values is greater than the preset threshold is greater than the stable threshold and is not greater than the global change threshold, the environment pattern classification module establishes a blank template in which the pixel points and the environment index distribution pattern are in one-to-one correspondence; recording the pixel points of which the average value of the absolute value of each difference value is greater than the preset threshold as current pixel points, and judging whether pixel points of which the average value of the absolute value of the difference value is greater than the preset threshold exist in the adjacent pixel points of the current pixel points or not; if yes, marking a blank template pixel point corresponding to the current pixel point position as 1; if not, marking the blank template pixel point corresponding to the current pixel point position as 0; for the pixel points of which the average value of the absolute value of the difference value is not more than the preset threshold value, marking the corresponding blank template pixel points as 0; after each pixel point of the blank template is marked in sequence, the blank template is converted into a binary pattern; counting the number of pixel points with the median value of 1 in the binary pattern, and judging whether the number of the pixel points is greater than a region change judgment threshold value or not; if the change is larger than the area change judgment threshold, the environmental index distribution pattern of the current data acquisition interval is classified into a local change environmental pattern, and if the change is not larger than the area change judgment threshold, the environmental index distribution pattern of the current data acquisition interval is classified into a global change environmental pattern;

the environment pattern feature extraction module divides the pattern into a plurality of blocks aiming at the condition that the environment index distribution pattern of the current data acquisition interval is classified as a stable environment pattern, calculates the average value of the values of all pixel points in each block, and extracts the average value of each block as the symbolic feature value of the pattern;

2. The cloud-computing-based machine room environment intelligent regulation system according to claim 1, wherein the cloud environment intelligent regulation platform comprises:

3. The intelligent regulation system for the computer room environment based on the cloud computing as claimed in claim 1, wherein the regulation and control strategy determination module preferentially calls a globally effective environment regulation facility for the case that the abnormal mode of the computer room environment is global abnormal; for the condition that the abnormal mode of the machine room environment is local abnormality, a local effective environment adjusting facility is preferentially called; for the condition that the abnormal mode of the machine room environment is identified as global change local abnormity, calling a local effective environment adjusting facility and simultaneously calling a global effective environment adjusting facility so as to avoid the local abnormity from diffusing along with the global change; for the condition that the abnormal mode of the machine room environment is local abnormal change, the adjusting power of the locally effective environment adjusting facility is further increased relative to the condition of the local abnormal so as to inhibit the local abnormal change trend; and if the abnormal mode of the machine room environment is global change abnormality, the adjusting power of the globally effective environment adjusting facility is further increased relative to the global abnormal condition so as to inhibit the global abnormal change trend.