WO2019109582A1

WO2019109582A1 - Non-intrusive energy consumption detection method and system

Info

Publication number: WO2019109582A1
Application number: PCT/CN2018/084250
Authority: WO
Inventors: 宁可; 彭富裕; 王春光
Original assignee: 亿可能源科技（上海）有限公司
Priority date: 2017-12-08
Filing date: 2018-04-24
Publication date: 2019-06-13
Also published as: CN109901442A; CN109901442B

Abstract

A non-intrusive energy consumption detection method and system, the non-intrusive energy consumption detection method comprising: implementing change point detection of an energy consumption sequence of an acquired load to obtain a change point sequence (S101); on the basis of pre-obtained energy consumption change information corresponding to a change in the operating state of a device, performing mapping recognition of the device operating state change event causing the change point sequence (S102) in order to perform energy consumption monitoring on the corresponding device in the load. By means of implementing change point detection of the collected energy consumption sequence of a load and on the basis of the pre-obtained energy consumption change information corresponding to the change in the operating state of the device, the present method determines the device operating state change event producing the change point, and thereby implements energy consumption detection of the device operating changes in a non-invasive manner.

Description

Non-intrusive energy consumption detection method and system

Technical field

The present application relates to the field of energy consumption monitoring and equipment automatic control, and in particular relates to a non-invasive energy consumption detecting method and system.

Background technique

In recent years, energy prices and consumption around the world have increased dramatically, and it is expected that this trend will continue to develop along the next few years. This is influenced by many factors, among which industrial, commercial and residential electricity consumption is growing rapidly with the improvement of people's ability to innovate, quality of life and consumption. According to statistics, electricity consumption is growing at a rate of 1.4% per year. Given the data and the continued construction of residential and commercial buildings, the continued rise in electricity consumption has raised awareness of energy efficiency.

In order to achieve energy-saving goals, appliances or equipment can be replaced by more energy-efficient alternatives, households or owners can change their behavior to reduce the use of energy-consuming equipment, and automated building management solutions can be used in building control equipment to achieve low energy consumption. Or arrange for operations that reduce energy costs during off-peak demand. However, these solutions all require the addition of sensor devices or even the modification of the device itself, which is detrimental to the operation of a more integrated management device to reduce the energy consumption of the device.

Summary of the invention

The present application provides a non-invasive energy consumption detection method and system for solving the problem of energy consumption detection for an invasively used device.

In order to achieve the above and other objects, the present application provides a method for detecting an energy consumption in a first aspect, comprising: performing a change point detection on a power consumption sequence of the acquired load to obtain a sequence of change points; and corresponding to a change in operating state of the device obtained in advance The energy consumption change information is mapped and identified to the device operating state change event sequence that causes the change point sequence to perform energy consumption monitoring on the corresponding device in the load.

In some embodiments of the first aspect, the method further comprises: obtaining the energy consumption sequence by a collection device coupled to the main power supply loop.

In some embodiments of the first aspect, the sampling interval of adjacent energy consumption data in the energy consumption sequence is in the range of (0, 60] s.

In some implementations of the first aspect, the manner of performing change point detection on the energy consumption sequence of the acquired load includes: using a Bayesian algorithm to change the energy consumption data in the energy consumption sequence Detection.

In some embodiments of the first aspect, the sequence of change points is a sequence of a plurality of energy subsequences comprising the detected change points acquired in the energy consumption sequence.

In some implementations of the first aspect, the method for mapping and identifying a sequence of device operating state change events that cause the sequence of changes is performed based on the energy consumption change information corresponding to the pre-determined device operating state change The method includes: mapping and identifying, according to the energy consumption change information and the energy consumption deviation of the energy consumption subsequence in the change point sequence, the change points of each change point in the change point sequence and each device operation state change event.

In some implementations of the first aspect, the method for mapping and identifying a sequence of device operating state change events that cause the sequence of changes is performed based on the energy consumption change information corresponding to the pre-determined device operating state change The method includes: adjusting, according to at least one of a pre-determined device energy consumption parameter and an influencing factor of the device operation, a mapping relationship between the change point in the change point sequence and a device running state change event.

In some embodiments of the first aspect, the method further includes an input interface that provides at least one of the device energy consumption parameter and the influencing factor of the device operation to obtain the corresponding device energy consumption parameter and the impact of the device operation. The steps of the factors.

In some embodiments of the first aspect, the device operating state change comprises at least one of a device start-stop state transition, and a transition between the plurality of operational states during operation of the device.

In some implementations of the first aspect, the energy consumption change information corresponding to the device operating state change is obtained in a non-intrusive manner.

In some embodiments of the first aspect, the method further comprises the step of machine learning the change point sequence to obtain the energy consumption change information.

In some embodiments of the first aspect, the device operating state change event includes a change in operating state of the frequency conversion device.

The second aspect provides a non-invasive energy consumption detecting system, comprising: a collecting device, configured to collect energy consumption data of the load; and at least one storage device, configured to save energy consumption data collected by the collecting device and At least one computer program; at least one processing device for executing the at least one computer program such that energy consumption detection of the method of any of the above is performed with an energy consumption sequence formed by the energy consumption data.

In some embodiments of the second aspect, the acquisition device collects energy consumption data over a range of (0, 60) s.

In some embodiments of the second aspect, the collection device is coupled to an energy meter coupled to the main power supply loop.

In some embodiments of the second aspect, the energy consumption data collected by the collection device includes energy consumption data generated by operation of the frequency conversion device.

In a third aspect, the application provides a server for communicating with a collecting device, wherein the collecting device is configured to collect energy consumption data of a load on a total power supply circuit, and the server includes: one or more processes One or more memories for storing energy consumption data provided by the collection device and at least one computer program; when the one or more computer programs are executed by the one or more processors, causing the The one or more processors perform energy consumption detection of the method of any of the above, using the energy consumption sequence formed by the energy consumption data.

In a fourth aspect, the present application provides a non-intrusive power consumption detection system, including: a change point detection module, configured to perform a change point sequence on a change point of an energy consumption sequence of the acquired load, and an energy consumption detection module, configured to And mapping, according to the energy consumption change information corresponding to the change of the operating state of the device, the device operating state change event sequence that causes the change sequence to perform energy consumption monitoring on the corresponding device in the load.

In some embodiments of the fourth aspect, the system further includes an energy consumption collection module for obtaining an energy consumption sequence sampled from the total power supply loop.

In some embodiments of the fourth aspect, the sampling interval of adjacent energy consumption data in the energy consumption sequence is in the range of (0, 60] s.

In some implementations of the fourth aspect, the change point detection module is configured to perform a change point detection on each energy consumption data in the energy consumption sequence by using a Bayesian algorithm.

In some embodiments of the fourth aspect, the change point sequence is a sequence of a plurality of energy subsequences including the detected change points acquired in the energy consumption sequence.

In some implementations of the fourth aspect, the energy consumption detecting module is configured to calculate, according to the energy consumption change information, an energy consumption deviation of the energy consumption subsequence in the change point sequence, in the change point sequence Each change point is mapped and identified with each device running state change event.

In some implementations of the fourth aspect, the energy consumption detecting module is configured to adjust a change point in the change point sequence based on at least one of a device energy consumption parameter obtained in advance and an impact factor of device operation The mapping relationship with the device running status change event.

In some embodiments of the fourth aspect, the system further includes an input module, configured to pre-acquire the input interface of at least one of the device energy consumption parameter and the influencing factor of the device operation At least one of equipment energy consumption parameters and factors affecting equipment operation.

In some embodiments of the fourth aspect, the device operating state change comprises at least one of a device start-stop state transition, and a transition between the plurality of operational states during operation of the device.

In some implementations of the fourth aspect, the energy consumption change information corresponding to the device operating state change is obtained in a non-intrusive manner.

In some embodiments of the fourth aspect, the system further comprises: a learning module, configured to perform machine learning using the sequence of changes to obtain the energy consumption change information.

In some embodiments of the fourth aspect, the device operating state change event comprises an operational state change of the frequency conversion device.

The non-invasive energy consumption detecting method and system provided by the present application determine the change by changing the energy consumption sequence of the collected load, and according to the energy consumption change information corresponding to the pre-determined operation state change of the device. The device running status change event of the point, thereby realizing the energy consumption detection of the device operation change of the concerned device in a non-invasive manner.

DRAWINGS

FIG. 1 is a flowchart of an embodiment of a non-intrusive energy consumption detecting method according to an embodiment of the present application.

2a, 2b, 2c, and 2d are schematic diagrams of energy consumption variation characteristic curves obtained by machine learning using a variable point sequence.

Figure 3 shows the energy consumption curve based on a sub-sequence of energy consumption.

4 is a schematic diagram of a device architecture of an embodiment of a non-intrusive power consumption detection system of the present application.

FIG. 5 is a schematic structural diagram of a device in another embodiment of the non-intrusive power consumption detecting system of the present application.

FIG. 6 is a schematic structural diagram of hardware included in a server of the present application in an embodiment.

7 is a schematic structural diagram of a program module of a non-intrusive energy consumption detecting system of an application in an embodiment

Detailed ways

The embodiments of the present application are described below by specific specific examples, and those skilled in the art can easily understand other advantages and effects of the present application by the contents disclosed in the present specification. The present invention may be embodied or applied in various other specific embodiments, and various modifications and changes may be made without departing from the spirit and scope of the invention.

It should be noted that the structures, the proportions, the sizes, and the like of the drawings in the present specification are only used to cooperate with the content disclosed in the specification for understanding and reading by those skilled in the art, and are not intended to limit the present application. The qualifications that can be implemented are not technically meaningful. Any modification of the structure, change of the proportional relationship or adjustment of the size should be continued without affecting the effects and objectives of the application. It is within the scope of the technical content disclosed in the present application. In the meantime, the terms "upper", "lower", "left", "right", "intermediate" and "one" as used in this specification are also for convenience of description, and are not intended to limit the present. The scope of the application can be implemented, and the change or adjustment of the relative relationship is considered to be within the scope of the application.

There are many types of electrical equipment in office parks, office buildings, industrial sites and residential buildings, such as electric lights, printers, computers and other small electrical equipment, as well as household appliances such as TVs and refrigerators, as well as elevators, central air conditioners, pumps, fans, etc. Large frequency conversion equipment (cluster). Among them, the use and deactivation of small appliances and household appliances are closely related to people's behavior and are not suitable for overall monitoring, while public equipment, especially large-scale inverter equipment (cluster), due to high power consumption and high concentration of industrial control, More intelligent monitoring of it will be beneficial to the efficient and energy-efficient operation of the equipment.

In order to coordinate and systematically control the operation of equipment in large places (especially large industrial frequency conversion equipment), it is impossible to install a control device on each equipment. To this end, in recent years, research on the use of non-invasive detection of energy consumption has gradually emerged. However, due to various factors such as large data volume, high data complexity, and various types of equipment, a large amount of data is required to perform machine learning on the operation process of key supervisory equipment in advance. This aspect requires the facility's equipment management system to cooperate with providing operational data specific to the supervised equipment for machine learning, and on the other hand to limit equipment updates within the premises, as each device update means that machine learning needs to be re-executed. In fact, the entire learning process, due to the large amount of data required, severely restricts the normal use of electrical equipment in the premises, and is not conducive to intelligent management of the operating equipment.

In order to solve the above problems and achieve energy consumption optimization monitoring purposes of a large power consumption device (cluster), the present application provides a non-intrusive energy consumption detection method. The non-intrusive energy consumption detection method is mainly performed by a non-intrusive energy consumption detection system. The non-intrusive energy consumption detection system includes software and hardware installed in a computer device. The computer device may be a server installed in the general control room of the above-mentioned place, a server (cluster) located in the network, or the like. The server (cluster) includes but is not limited to: a single server, a distributed server, and a cloud service platform.

The cloud service platform includes a public cloud (Public Cloud) service platform and a private cloud (Private Cloud) service platform, where the public or private cloud service platform includes Software-as-a-Service (software as a service, referred to as SaaS), Platform-as-a-Service (Platform as a Service, abbreviated as PaaS) and Infrastructure-as-a-Service (Infrastructure as a Service, IaaS for short). The private cloud service platform is, for example, an Alibaba Cloud computing service platform, an Amazon cloud computing service platform, a Baidu cloud computing platform, a Tencent cloud computing platform, or an intra-enterprise private cloud.

The data used by the non-intrusive energy consumption detection system for energy consumption detection is derived from a sequence of power consumption data collected online by the monitored location, that is, an energy consumption sequence. To this end, the non-invasive energy consumption detection system further includes an acquisition device capable of acquiring the energy consumption sequence and software built in the corresponding device. The energy consumption sequence provides a plurality of energy consumption data arranged in a sampling order. The energy consumption data is exemplified but not limited to: power data, current data, voltage data, power factor, and the like. Wherein, the collecting device may be a special collecting device connected to the total power supply circuit of the monitored place or an electric energy meter with built-in energy consumption data sampling function. For example, a dedicated collection device is coupled to an energy meter mounted on the power supply loop and continuously collects energy consumption data to form an energy consumption sequence. Alternatively, the collection device is a third-party device capable of providing a power consumption sequence, such as a third-party energy meter with built-in energy consumption data sampling function, and a third-party power supply monitoring system. The power monitoring system can perform power supply monitoring on devices connected to each circuit loop by analyzing the collected energy consumption sequence. For example, the non-intrusive energy consumption detection system acquires an energy consumption sequence from a power supply control system of the factory. For another example, the non-invasive energy consumption detection system acquires an energy consumption sequence from a remote power supply monitoring system of the office building group.

Here, the sampling interval of the energy consumption sequence may be determined by a sampling device that provides the energy consumption sequence. In some specific examples, a dedicated acquisition device directly connected to the energy meter or an energy meter with a built-in acquisition function can collect energy consumption data at preset sampling intervals. Wherein, the bandwidth of the metering device such as the connected energy meter is limited, or the bandwidth of the data channel (such as mobile data network, local area network, ad hoc network, Internet of Things, power network, etc.) accessed by the collecting device is output. The limitation of the data amount of the data, or the hardware capacity of the metering device such as the connected power meter and the bandwidth of the data path accessed by the collecting device, the sampling interval is (0, 60] The data volume requirement of the energy consumption detection of the present application is more satisfied in the range. For example, the sampling interval is 1 s, 2 s, 3 s, 4 s, 5 s, 6 s, 7 s, 8 s, 9 s, 10 s, 11 s, 12 s, 13 s. , 14s, 15s, 16s, 17s, 18s, 19s, 20s, 21s, 22s, 23s, 24s, 25s, 26s, 27s, 28s, 29s, 30s, 31s, 32s, 33s, 34s, 35s, 36s, 37s, 38s , 39s, 40s, 41s, 42s, 43s, 44s, 45s, 46s, 47s, 48s, 49s, 50s, 51s, 52s, 53s, 54s, 55s, 56s, 57s, 58s, 59s, or 60s. The sampling interval is a time interval with a time interval accuracy greater than 0.1 s, or even a higher time interval precision, such as 1.1 s, 1 .11s, etc. It should be noted that the above sampling interval is only a exemplifying time interval for sampling the energy consumption data, instead of indicating that the sampling interval uses only the above examples to perform sampling processing of energy consumption data. With the hardware improvement of the electric energy meter or the reduction of the data transmission cost, the sampling interval can be reduced to within 1 s, such as 0.5 s, etc., and the non-invasive energy consumption detecting system can provide more detailed energy according to the sampling device. The consumption sequence performs more accurate energy consumption detection. In other words, those skilled in the art, in the context of knowing the innovative spirit of the present application, should also adjust the above-mentioned time interval according to the actual power meter performance and the application of the data transmission technology. The scope covered by the application.

Wherein, the acquisition device used in the application may be provided with a power sensor or a voltage sensor for measuring the energy consumption data on the load side, or the collection device reads the energy consumption data from the database and transmits it through the data line or the network. Computer equipment for energy consumption testing. The acquisition device can transmit the collected energy consumption data to the computer device synchronously or asynchronously.

The non-intrusive energy consumption detection system performs the energy consumption analysis on the acquired energy consumption sequence to obtain a mapping relationship between the operating state change event and the energy consumption change of the monitored device, and the mapping relationship is The monitored device performs energy consumption monitoring. The energy consumption monitoring includes, but is not limited to, at least one of: detecting faults of the monitored equipment, measuring power consumption of the monitored equipment, and optimizing operation of the monitored equipment to reduce energy consumption.

Here, the monitored device is usually a device that is of high concern for the site management or the property owner, such as a frequency conversion device, a large power consumption device, etc., but is not limited thereto, and can be designed according to actual needs by those skilled in the art. The energy consumption detection scheme described in the present application performs energy consumption detection on the operating state events of the fixed frequency device and the small power consumption device. Here, the high power consumption device is exemplified by a device having a rated power or a maximum power of more than kilowatts, or a device having a rated power and a maximum power of more than kilowatts. The low-power devices are exemplified by devices with rated power or maximum power below kilowatts, or devices with rated power and maximum power below kilowatts.

Here, in relation to the number of devices to be supervised and the number of operating states that can be changed by a single device, the device running change event may be attributed to all operating states of one or more devices, wherein the types of the devices may be the same or different. The device operating state change includes at least one of a device start-stop state transition and a transition between the plurality of operational states during operation of the device. For example, the monitored equipment includes: three cold water pumps, two drive motors, and two heating devices; wherein, the operating state changes of the cold water pump include but are not limited to: state change between start-stop, standby-operating state The change between the water cycle speed, the change of the water cycle speed state, the working state of each gear shift to the stop state, etc.; the running change properties of the drive motor include but are not limited to: state change between start-stop, standby-operating state The change between the drives, the change of the drive gear state, the working state of each gear shift to the stop state, etc.; the operational change attributes of the heating device include but are not limited to: state change between start-stop, standby-work state The change, the change between the state of the heating gear, the working state of each gear shift to the stop state, and the like. The moment when the energy consumption changes due to the change of the operating state is called the change point.

It can be seen that a change point may be caused by a change in the operating state of one device or a combination of operating states of multiple devices. We refer to a combination of a running state change or a change in the operating state of multiple devices that can cause a change point as a device operating state change event. In order to more accurately monitor the energy consumption of the devices of interest, we are more concerned with the device operating state change events caused by the changes in the operating state of these devices. Wherein, when the monitored device includes the variable frequency device, the device operating state change event includes a change in the operating state of the variable frequency device. Please refer to FIG. 1 , which shows a flowchart of an embodiment of the non-intrusive energy consumption detecting method of the present application in an embodiment. As shown, the non-intrusive energy consumption detection system performs the following steps:

In step S110, a change point detection is performed on the energy consumption sequence of the acquired load to obtain a change point sequence. The load includes all the loads on the loop where the energy consumption sequence is collected, including but not limited to: small appliances such as lamps, computers, printers, household appliances such as refrigerators and televisions, and monitored inverters such as elevators and central air conditioners. Or fixed frequency equipment, etc.

Since the solution of the present application adopts non-invasive energy consumption detection, the consumed energy sequence reflects the power consumption of all loads on the collected power supply line along the time axis. When there is equipment on the load side from stop to start, from run to stop, gear position adjustment during operation, and switching between standby state and working state, the load side energy consumption will change. Obviously, when the monitored operating state of the device changes, its corresponding energy consumption change will also be reflected in the acquired energy consumption sequence.

By performing a change point detection on the obtained energy consumption sequence, a time change of each load energy consumption is obtained, wherein the change time of the load energy consumption (ie, the change point) includes energy caused by a change in the operating state of the monitored device. Time to change.

Here, the step S110 can select a corresponding change point detection mode according to the data type of the energy consumption sequence. For example, if the energy consumption sequence is a voltage data type, the change point detection may be performed by using a harmonic analysis method or the like. If the energy consumption sequence is a power data type, the change point detection may be performed by using Allen variance and load detection. In some embodiments, the step S110 performs a change point detection on each energy consumption data in the energy consumption sequence by using a Bayesian algorithm. For example, detecting the first energy consumption data with the initial condition and the posterior condition is the possibility of sampling before the change point. When the probability meets the sampling probability threshold before the change point, the first energy consumption of the sampling is determined. The time corresponding to the data is before the change point; then, according to the determined result of the first energy consumption data corresponding to the change point, the a priori condition and the posterior condition are adjusted and the second energy consumption data change point is detected The possibility of energy consumption data, when the possibility meets the pre-change condition, it is determined that the time corresponding to the second energy consumption data is still sampled before the change point. The energy consumption data is iteratively detected until it is detected that the sampling time corresponding to an energy consumption data is after the change point, and then the change point is determined to occur within the sampling interval of the energy consumption data before and after the change point. The Bayesian algorithm is used to detect a change point in the energy consumption sequence.

After the change of the energy consumption sequence is detected, the energy consumption data before and after each change point can be extracted from the energy consumption sequence, and the change point sequence is formed in time sequence. Or the change sequence of the energy consumption data corresponding to each change point forms a sequence of change points in time sequence. In some embodiments, the sequence of change points is a sequence of a plurality of energy subsequences comprising the detected change points acquired in the energy consumption sequence. For example, the obtained energy consumption sequence is {p ₁ , p ₂ , p ₃ , ... p _n }, and after the change point detection, the i energy consumption data before the change point in the energy consumption sequence and The j energy consumption data after the change point forms a power consumption subsequence corresponding to the change point; the energy consumption subsequences are arranged in time sequence to form a change point sequence, and step S120 is performed. Where n and j are both positive integers, i≥0, where i and j may be equal or unequal.

In step S120, based on the energy consumption change information corresponding to the device operating state change obtained in advance, the device operating state change event sequence causing the change point sequence is mapped and identified to perform energy consumption on the corresponding device in the load. monitor.

The energy consumption change information corresponding to the operating state change of the device includes, but is not limited to, a power consumption change threshold range, an energy consumption change curve, and the like. Among them, since the frequency conversion technology is adopted during each state transition of the frequency conversion device, each state transition may correspond to one or more energy consumption change modes. Therefore, a device operating state change may correspond to only one energy consumption change information or corresponding multiple energy consumption change information. In order not to interfere with the normal operation of the equipment in the monitored area, these energy consumption change information is obtained in a non-intrusive manner. For example, it can be learned by pre-simulating the change of the operating state of the device or calculated according to the electrical parameters of the device.

Since the actual energy consumption sequence can provide more realistic and complex energy consumption noise, and the above-mentioned energy consumption change information is determined to be idealized, the accuracy of establishing the above mapping relationship needs to be improved. To this end, in some embodiments, the energy consumption change information is obtained by machine learning. Here, since the present application uses non-intrusive energy consumption detection, the method described in the present application further includes a step S130 (not shown) that is executed before the execution of the step S120 or in parallel with the step S120.

In step S130, machine learning is performed by using the change point sequence to obtain the energy consumption change information. In some embodiments, the energy subsequence corresponding to each change point is obtained by accumulating for a period of time; according to the power consumption parameter of the monitored device, the electrical characteristics of the main electrical device in the device, etc., characteristic analysis of each energy subsequence is performed to obtain a device. The change in energy consumption corresponding to the change in operating status. In other embodiments, the energy subsequence corresponding to each change point is obtained by accumulating for a period of time; each energy subsequence is clustered, and then the energy consumption subsequence of the same classification is analyzed to obtain energy consumption change information; The power consumption parameter of the monitoring device and the electrical characteristics of the main electrical device in the device are matched, and the change of the operating state of the device is matched with the energy consumption change information of each category, so that the energy consumption change information corresponding to the change of the operating state of the device is obtained.

The obtained energy consumption change information includes at least one of the following: an energy consumption change trend curve, an energy consumption mean change curve, and an energy consumption change threshold range. For example, please refer to FIG. 2a, 2b, 2c, 2d, which is a schematic diagram showing a characteristic curve of energy consumption obtained by machine learning using a sequence of changing points.

It should be noted that the step S130 can perform continuous machine learning according to the obtained sequence of change points, and continuously update the obtained energy consumption change information, so that step S120 can be provided with more accurate energy consumption change information to improve Detection accuracy of non-invasive energy detection systems.

After obtaining the energy consumption change information corresponding to the running state change of the device, the non-intrusive energy consumption detecting system is based on the energy consumption change information corresponding to the running state change of the device, and the device running state change event that causes the change point sequence The sequence is mapped and identified.

Here, according to the monitored device and the number of changes in its operating state, a corresponding number of device operating state change events are preset; the energy subsequence corresponding to each change point in the change point sequence, energy consumption data or energy consumption The data change value is matched with each energy consumption change information, and each matching degree reflects the possibility P1 of each operation state change of each device at the change point time; and the device operation state change event generated at a single change point is determined based on each possibility P1. Possibility P2, and the possibility P3 associated with the combination of device operating state change events of a plurality of change points in the sequence of change points; the probability P2 corresponding to each device operating state change event based on the change point sequence And P3, obtaining a mapping relationship between each change point in the change point sequence and each device running state change event. Therefore, the non-intrusive energy consumption detecting system can detect the energy consumption change when the operating state of the corresponding device changes according to the obtained mapping relationship.

Among them, different state transitions of the same device may be mutually exclusive. For example, a state transition from start to stop and a state transition from stop to start are not possible at the same time. In addition, during the operation of multiple devices based on timing generation, different state transitions of the same device and state transitions between different devices are mutually exclusive. For example, if the change point A1 is caused by the device D1 transitioning from the start to the stop state, the internal change point of this interval may not be initiated by the device D1 until the change point of the device D1 transition from the stop to the start state is not detected. Any state transition between a stop state transition, or device D1 operation or standby. Therefore, when the initial assignments are possible for each possibility P1, P2 and P3, when calculating the possibilities P1, P2 and P3, or when determining the mapping relationship based on the possibilities P2 and P3, based on The above exclusion condition is adjusted according to the corresponding possibility or the mapping relationship, or the corresponding possibility and the mapping relationship are adjusted based on the above exclusion condition to improve the detection accuracy.

In some embodiments, based on the energy consumption change information and the energy consumption deviation of the energy consumption subsequence in the change point sequence, each change point in the change point sequence is mapped and identified with each device operation state change event. Specifically, the non-intrusive energy consumption detection system performs the following steps:

In step S121, traversely compare all the energy consumption change information with a single energy subsequence in the change point sequence to obtain a set of energy consumption deviations, and repeat the above traversal step to obtain all the energy subsequences in the change point sequence. The energy consumption deviation group. For example, please refer to FIG. 2a, 2b, 2c, 2d and FIG. 3, wherein each of the energy consumption change information E1-E4 in FIG. 2a, 2b, 2c, 2d takes the following correspondence as an example: the energy consumption change information E1 is described. For the state change of the device A1 from start to stop, the energy consumption change information E2 is described as a state change of the device A1 from stop to start, and the power consumption change information E3 is described as a state change of the device A2 from start to stop, and energy consumption. The change information E4 is described as a state change of the device A2 from stop to start; FIG. 3 shows a change curve of the energy consumption based on a power subsequence. The energy consumption variation curve of the energy subsequence L1 and the energy consumption change information E1-E4 are calculated one by one to obtain four energy consumption deviations, and the four energy consumption deviations are used as one energy consumption deviation group. The manner of calculating the deviation includes, but is not limited to, any one or more of the following: calculating a power consumption difference curve, calculating an average value of the energy consumption difference, calculating a characteristic error of the energy consumption change, and the like. According to the above deviation calculation method, the respective energy consumption deviation groups of the corresponding energy sub-sequences L2, ..., Lm are obtained.

In step S122, based on the obtained energy consumption deviation group, the possibility that the corresponding change point generates each device operating state change event is calculated. For example, the preset statistical algorithm is used to calculate the possibility of the change of the operating state of the device corresponding to each energy consumption deviation in the single energy consumption deviation group, and then the possibilities obtained by adjusting the mutually exclusive conditions between the operating change events of the devices are adjusted. In turn, the probability of each group of energy consumption deviations corresponding to each device operating state change event is determined. The statistical algorithm includes but is not limited to: a Viterbi algorithm, a Bayesian algorithm, and the like.

In step S123, the mapping relationship between the change point sequence and the device operating state change event sequence is determined by generally evaluating the possibility of each change point and each device running a change event. For example, using a statistical algorithm such as maximum likelihood estimation or least squares estimation to estimate the probability of mapping between each change point and the device operation change event, and selecting the estimation result to determine the optimal possibility to determine each change point. The mapping relationship between the change sequence and the device running state change event sequence is obtained.

Due to the complex interference factors such as noise in the energy subsequence, unpredicted load equipment, and the error between the obtained energy consumption change information and the actual energy consumption change information, we introduced energy consumption detection and monitored equipment. Relevant information, intervening on the mapping relationship, thereby improving the accuracy of the mapping relationship.

In an implementation manner, the step S120 further includes the step S124: adjusting the change point of the change point sequence and the running state of the device based on at least one of the pre-determined device energy consumption parameter and the influencing factor of the device operation. The mapping relationship of events. The energy consumption parameters of the device are exemplified by the rated power, the maximum power, and the like of the device, and other parameters related to the energy consumption of the device. The energy consumption parameter of the device may be obtained from a product specification, a specification, a user manual of the device, or may be provided by a park management party or a property management party. The influencing factors include, but are not limited to, weather, occupancy rate, equipment start-stop schedule, etc. These influencing factors may directly or indirectly become factors in the change of the operating state of the equipment. For example, if the weather is lower than t1, the maintenance personnel of the equipment turn on the heating equipment. The weather maintenance equipment only installs the fresh air equipment between t1 and t2 degrees. The equipment maintenance personnel turn on the cold water pump and the fan equipment when the weather is above t2 degrees. For another example, the equipment maintenance personnel control the start and stop of each equipment of the building air conditioning system according to the equipment start-stop schedule. The influencing factors can be obtained through a network from a third party service device, through a manually imported configuration file, or through a human machine interface.

In order to make the non-invasive energy consumption detecting system more suitable for a plurality of places, a specific example is that an input interface of at least one of the device energy consumption parameter and the influencing factor of the device operation is provided in advance to obtain the corresponding device. Energy consumption parameters and factors affecting equipment operation. For example, the monitoring platform provided to the user includes an input interface that can input the energy consumption parameters of the monitored devices and the influencing factors of the device operation. When the user clicks the submit button, the non-invasive energy consumption detecting system acquires the Equipment energy consumption parameters in the interface and factors affecting equipment operation. The non-intrusive energy consumption detection system assigns the acquired energy consumption parameters of the equipment and the influencing factors of the equipment operation to corresponding parameters in the preset statistical algorithm.

Since the energy consumption parameters of the equipment and the influencing factors of the equipment operation may affect the operation state of the equipment, it may optimize multiple steps of the mapping relationship determination process. For example, according to the weather information in the obtained influencing factors, the weather is less than 10 degrees, and the non-invasive energy consumption detecting system may only calculate the energy deviation of the energy subsequence corresponding to each change point and the operating state change of the heating device. For example, according to the acquired influencing factors, the equipment running schedule is to start the cold water pump at 8:00 am on Monday, and the non-invasive energy consumption detection system can improve the change point generated in the 7:50-8:10 period and contain cold. The possibility of mapping the event of the pump from the stop to the start state transition, or increasing the transition of the cold water pump from stop to start state in each event, or including the transition from the stop to the start state of the cold water pump, or both of the above The weight of the evaluation of the event. For another example, according to the rated power in the energy consumption parameter of the device, the error between the obtained energy consumption deviation and the rated power is evaluated. If the error is determined to be within the preset reliability range, the corresponding change point may be improved. Possibilities and/or weights caused by changes in the operating state of the respective device.

It should be noted that the foregoing improvement possibility and the weight increase are relative to the possibility and weight of the device operating state change that are not affected by the device energy consumption parameter and the influence factor of the device operation. Those skilled in the art can increase the likelihood and weight associated with changes in the operational state of the affected equipment by reducing the likelihood and weight of unaffected or very influential operating state changes.

The present application considers the influence of the obtained device energy consumption parameter and the influencing factors of the device operation on the mapping relationship between the change point sequence and the device operating state change event sequence, and the device operating state change event sequence of the change point sequence. The mapping relationship between the two is adjusted to obtain a more accurate correspondence between the mappings. Therefore, it is possible to provide monitoring information for equipment maintenance personnel, property management parties, and the like who use the corresponding equipment and care about the operation of the corresponding equipment. For example, the energy consumption detection determines that the sequence of equipment operation change events corresponding to the sequence of change points includes: three cold water pumps are in a startup state, and one fan device is in a startup state, and a cold water pump is started according to the obtained sequence of events. Too many, you can turn off the prompts for two of them. For example, the energy consumption detection determines that the sequence of equipment operation change events corresponding to the change point sequence includes: the energy consumption of the cold water pump from stop to start during 7:00-9:00, and the standby and work of the cold water pump According to the sequence of events, a fault prompt corresponding to the cold working time of the cold water pump can be obtained.

In addition, the present application also provides a non-invasive energy consumption detection system. Please refer to FIG. 4 and FIG. 5 , which respectively show the architecture of the device in a different embodiment of the non-intrusive energy consumption detection system of the present application. The non-invasive energy consumption detection system includes an acquisition device 11, at least one storage device 12, and at least one processing device 13.

The collecting device 11 is configured to collect energy consumption data of the load 21 . The load 21 is a powered device connected to the total power supply circuit 22. The collection device 11 can be a dedicated collection device 11 connected to the total power supply circuit 22 of the monitored location or an energy meter with built-in energy consumption data sampling function. For example, the dedicated collection device 11 is coupled to an energy meter mounted on the power supply circuit 22 and continuously collects energy consumption data to form an energy consumption sequence. Alternatively, the collection device 11 is a third-party device capable of providing an energy consumption sequence, such as a third-party energy meter with built-in energy consumption data sampling function, a third-party power supply monitoring system, and the like. The power monitoring system can perform power supply monitoring on devices connected to each circuit loop by analyzing the collected energy consumption sequence. For example, the non-intrusive energy consumption detection system acquires an energy consumption sequence from a power supply control system of the factory. For another example, the non-invasive energy consumption detection system acquires an energy consumption sequence from a remote power supply monitoring system of the office building group. Since the energy consumption data collected by the sampling device is derived from the energy consumption data of the load 21 on the total power supply circuit 22, when the load 21 includes the frequency conversion device, the energy consumption data collected by the collection device 11 includes the frequency conversion device. Run the generated energy consumption data. The frequency conversion device includes but is not limited to: a large-scale frequency conversion device (cluster), a small-scale frequency conversion device, and the like. For example, the large-scale frequency conversion equipment (cluster) such as an elevator, a central air conditioner, an industrial machine, and the like, a common frequency conversion device; the small-scale frequency conversion device such as a refrigerator, a home air conditioner, and the like. Additionally, the load 21 may include high power devices and low power devices, depending on power. The high-power devices are exemplified by rated power, or devices with a maximum power of more than kilowatts, or devices with rated power and maximum power above kilowatts. The low power consumption device is exemplified by a device having a rated power or a maximum power of less than kilowatts, or a device having a rated power and a maximum power of less than kilowatts.

Here, the sampling interval of the energy consumption sequence may be determined by a sampling device that provides the energy consumption sequence. In some specific examples, the dedicated collection device 11 directly connected to the energy meter or the energy meter with built-in acquisition function may collect energy consumption data according to a preset sampling interval. Wherein, the bandwidth capability of the metering device such as the connected energy meter is limited, or the bandwidth of the data channel (such as mobile data network, local area network, ad hoc network, Internet of Things, power network, etc.) accessed by the collecting device 11 The limitation of the amount of data consumed by the data, or the hardware capacity of the metering device such as the connected power meter and the bandwidth of the data path accessed by the collecting device, the sampling interval is (0) , 60] can better meet the data volume requirement of the application for energy consumption detection. For example, the sampling interval is 1s, 2s, 3s, 4s, 5s, 6s, 7s, 8s, 9s, 10s, 11s, 12s, 13s, 14s, 15s, 16s, 17s, 18s, 19s, 20s, 21s, 22s, 23s, 24s, 25s, 26s, 27s, 28s, 29s, 30s, 31s, 32s, 33s, 34s, 35s, 36s, 37s, 38s, 39s, 40s, 41s, 42s, 43s, 44s, 45s, 46s, 47s, 48s, 49s, 50s, 51s, 52s, 53s, 54s, 55s, 56s, 57s, 58s, 59s, or 60s. The sampling interval is a time interval with a time interval precision greater than 0.1 s, or even a higher time interval precision, such as 1.1 s, 1 .11s, etc. It should be noted that the above sampling interval is only a exemplifying time interval for sampling the energy consumption data, instead of indicating that the sampling interval uses only the above examples to perform sampling processing of energy consumption data. With the hardware improvement of the electric energy meter or the reduction of the data transmission cost, the sampling interval can be reduced to within 1 s, such as 0.5 s, etc., and the non-invasive energy consumption detecting system can provide more detailed energy according to the sampling device. The consumption sequence performs more accurate energy consumption detection. In other words, those skilled in the art, in the context of knowing the innovative spirit of the present application, should also adjust the above-mentioned time interval according to the actual power meter performance and the application of the data transmission technology. The scope covered by the application.

Wherein, the collecting device 11 used in the present application may be provided with a power sensor or a voltage sensor for measuring the energy consumption data on the load 21 side, or the collecting device 11 reading the energy consumption data from the database and passing through the data line or the network. Transfer to a computer device for energy consumption detection. The energy consumption data is exemplified but not limited to: power data, current data, voltage data, and the like. The acquisition device 11 can transmit the collected energy consumption data to the computer device synchronously or asynchronously. The energy consumption data is stored by the computer device in at least one storage device 12.

The storage device 12 can be integrated with the collection device 11, for example, the storage device 12 is part of a server (group) where the power monitoring system is located, or the storage device 12 is located in a remote server (group).

Here, the storage device 12 may include a high speed random access memory, and may also include a non-volatile memory, such as one or more magnetic disk storage devices, flash memory devices, or other non-volatile solid state storage devices. In some embodiments, storage device 12 may also include memory remote from one or more processors, such as network attached storage accessed via RF circuitry or external ports and a communication network (not shown), wherein the communication The network may be the Internet, one or more intranets, a local area network (LAN), a wide area network (WLAN), a storage area network (SAN), etc., or a suitable combination thereof. The memory controller can control access to the storage device 12 by other components of the device, such as the CPU and peripheral interfaces.

The storage device 12 can be provided by one storage device or distributed among multiple storage devices. For example, depending on the actual area in which the monitored location is located (e.g., Haidian District, Beijing), the storage device 12 can be placed nearby. As another example, the storage device 12 is a storage server for providing a cloud service platform.

The storage device 12 may allow one or more processing devices 13 to perform energy consumption data read and write operations. The collection device 11 stores the energy consumption data in the corresponding storage device 12 by the read and write operations of the processing device 13. At least one computer program is also stored in at least one of said storage devices 12. The computer program is used to be executed by the processing device 13.

The processing device 13 may be configured in the same computer device as the storage device 12, or configured in different computer devices, to execute the at least one computer program, so that the energy formed by the energy consumption data The consumption sequence performs energy consumption detection based on the non-invasive energy consumption detection method described in the present application. The computer device may be a single server, a component device of a server cluster, or a server device in a cloud service platform. The computer device can be placed in the total control room of the monitored location, or in a third-party server storage room, or other physical location capable of communicating with the collection device 11 and reading and writing data with the storage device 12.

The working process of the non-intrusive energy consumption detecting system is briefly described by using FIG. 3 as an example: the collecting device 11 collects the energy consumption data of the load 21 on the total power supply circuit 22 of the monitored location according to a preset sampling interval, and transmits the data to the processing. The device 13, the processing device 13 saves it in the storage device 12. The energy consumption sequence formed by the energy consumption data arranged in the order of sampling timing is used by the processing device 13 for non-intrusive machine learning for a predetermined period of time. Specifically, the processing device 13 performs a change point detection on the acquired energy consumption sequence by using a Bayesian algorithm, and performs clustering and feature analysis based on a curve formed by the energy subsequence including the change point, thereby obtaining a desired Monitors energy consumption change information corresponding to changes in the operating status of the device. The duration of the machine learning and the detection efficiency of the energy consumption detection are both taken into consideration, and the energy consumption change information is still performed during the energy consumption detection, and the energy consumption change information corresponding to the change of the operating state of each device is updated in a timely manner. In order to improve the accuracy of the energy consumption detection, the processing device 13 further provides an input interface for the maintenance personnel running by the management site device to fill in the device energy consumption parameters and influencing factors related to the monitored device operating state changes. During the energy consumption detection, the processing device 13 performs error calculation on each of the detected energy consumption sub-sequences and the energy consumption change information in the vicinity of each change point, and obtains energy consumption deviations of each group corresponding to each change point; The group energy consumption deviation initially determines the possibility of generating a device operating state change event for each change point; and further adjusts each of the possibilities based on the device energy consumption parameter and the influencing factor to obtain the most likely to conform to the facts. The mapping relationship between the point and each device running state change event.

The application also provides a server. Please refer to FIG. 6, which is a schematic diagram of the hardware structure included in the server. As shown, the server 3 includes one or more memories 31 and at least one processor 32. The server 3 can be configured on a single server, a server cluster, or a cloud service platform. The server cluster includes, but is not limited to, a distributed server group. The cloud service platform includes a public cloud service platform and a private cloud service platform, where the public or private cloud service platform includes Software-as-a-Service (SaaS). , Platform-as-a-Service (Platform as a Service, abbreviated as PaaS) and Infrastructure-as-a-Service (Infrastructure as a Service, referred to as IaaS). The private cloud service platform is, for example, an Alibaba Cloud computing service platform, an Amazon cloud computing service platform, a Baidu cloud computing platform, a Tencent cloud computing platform, or an intra-enterprise private cloud.

Correspondingly, the memory 31 and the processor 32 may be configured in a single server or in multiple servers. One memory 31 and a plurality of processors 32 can be configured in each server. The server may be placed in the vicinity of the actual area where the monitored location is located (such as Haidian District, Beijing), or in the self-use machine room or the third-party computer room according to the cloud service platform or the device management needs of the server cluster.

The memory 31 can include high speed random access memory and can also include non-volatile memory, such as one or more magnetic disk storage devices, flash memory devices, or other non-volatile solid state storage devices. In some embodiments, the storage device may also include a memory 31 remote from one or more processors 32, such as network attached storage accessed via RF circuitry or external ports and a communication network (not shown), wherein the communication network It can be the Internet, one or more intranets, a local area network (LAN), a wide area network (WLAN), a storage area network (SAN), etc., or a suitable combination thereof. The memory controller can control access to the storage device by other components of the device, such as the CPU and peripheral interfaces.

The processor 32 includes a single core or multi-core processor. The processor 32 is operatively coupled to the memory 31 and/or the non-volatile storage device. More specifically, processor 32 may execute instructions stored in a memory and/or non-volatile storage device to perform operations in a computing device, such as generating image data and/or transmitting image data to an electronic display. As such, processor 32 may include one or more general purpose microprocessors, one or more application specific processors (ASICs), one or more field programmable logic arrays (FPGAs), or any combination thereof.

The processor 32 is also operatively coupled to a network interface to communicatively couple the computing device to the network. For example, the network interface can connect the computing device to a personal area network (PAN) (such as a Bluetooth network), a local area network (LAN) (such as an 802.11x Wi-Fi network), and/or a wide area network (WAN) (inject a 4G or LTE cellular network) . Through the network interface, the processor 32 can acquire the energy consumption data provided by the collection device 4 and store the energy consumption data into the memory 31; even, the processor 32 can acquire a third party or Equipment energy consumption parameters and influencing factors provided by users (site equipment maintenance personnel).

The memory 31 also stores at least one computer program, and the processor 32 can call and execute the computer program to perform non-invasive energy consumption detection based on the energy consumption sequence formed by the energy consumption data. The energy consumption test provided by the method.

FIG. 6 is a schematic diagram illustrating the working process of the server. The server acquires and saves the energy consumption data of the load on the total power supply loop of the monitored location according to a preset sampling interval. The energy consumption sequence formed by the server in accordance with the energy consumption data arranged in the order of sampling timing is used for non-intrusive machine learning in a predetermined period of time. Specifically, the server uses a Bayesian algorithm to perform a change point detection on the acquired energy consumption sequence, and performs clustering and feature analysis based on a curve formed by the energy subsequence including the change point, thereby obtaining desired monitoring. The change in energy consumption corresponding to the change in the operating state of the device. The duration of the machine learning and the detection efficiency of the energy consumption detection are both taken into consideration, and the energy consumption change information is still performed during the energy consumption detection, and the energy consumption change information corresponding to the change of the operating state of each device is updated in a timely manner. In order to improve the accuracy of the energy consumption detection, the server also provides an input interface for the maintenance personnel running by the management site equipment to fill in the equipment energy consumption parameters and influencing factors related to the monitored equipment operating state changes. During the energy consumption detection, the processing device performs error calculation on each of the detected energy subsequences and energy consumption change information in the vicinity of each change point, and obtains energy consumption deviations of each group corresponding to each change point; The energy consumption deviation initially determines the possibility of generating a device operating state change event for each change point; and further adjusts each of the possibilities based on the device energy consumption parameter and the influencing factor to obtain each change point that most likely meets the fact The mapping relationship with each device running status change event.

The application also provides a non-invasive energy consumption detection system. The non-intrusive energy consumption detection system includes software and hardware installed on a computer device. The computer device may be as described above with respect to the computer device or based on the computer device described above, and will not be described in detail herein.

The non-intrusive energy consumption detection system includes program modules capable of controlling hardware in a computer device to execute in time series. Please refer to FIG. 7 , which is a schematic structural diagram of a program module of the non-intrusive energy consumption detecting system of the present application. As shown in the figure, the non-invasive energy detecting system includes at least a change point detecting module 51 and an energy detecting module. 52 and energy consumption collection module 53.

Here, the data used by the non-intrusive energy consumption detecting system 5 for energy consumption detection is derived from the power consumption data sequence collected online by the monitored location, that is, the energy consumption sequence. To this end, the non-invasive energy consumption detection system 5 acquires energy consumption data from the acquisition device. The energy consumption sequence provides a plurality of energy consumption data arranged in a sampling order. The energy consumption data is exemplified but not limited to: power data, current data, voltage data, power factor, and the like.

Wherein, the collecting device may be a special collecting device connected to the total power supply circuit of the monitored place or an electric energy meter with built-in energy consumption data sampling function. For example, a dedicated collection device is coupled to an energy meter mounted on the power supply loop and continuously collects energy consumption data to form an energy consumption sequence. Alternatively, the collection device is a third-party device capable of providing a power consumption sequence, such as a third-party energy meter with built-in energy consumption data sampling function, and a third-party power supply monitoring system. The power monitoring system can perform power supply monitoring on devices connected to each circuit loop by analyzing the collected energy consumption sequence. For example, the non-intrusive energy consumption detection system 5 obtains an energy consumption sequence from a power supply control system of the factory. For another example, the non-intrusive energy consumption detection system 5 obtains an energy consumption sequence from a remote power supply monitoring system of the office building group.

Here, the sampling interval of the energy consumption sequence may be determined by a sampling device that provides the energy consumption sequence. In some specific examples, a dedicated collection device directly connected to the energy meter or an energy meter with a built-in acquisition function may collect energy consumption data at preset sampling intervals. Wherein, the bandwidth of the metering device such as the connected energy meter is limited, or the bandwidth of the data channel (such as mobile data network, local area network, ad hoc network, Internet of Things, power network, etc.) accessed by the collecting device is output. The limitation of the data amount of the data, or the hardware capacity of the metering device such as the connected power meter and the bandwidth of the data path accessed by the collecting device, the sampling interval is (0, 60] The data volume requirement of the energy consumption detection of the present application is more satisfied in the range. For example, the sampling interval is 1 s, 2 s, 3 s, 4 s, 5 s, 6 s, 7 s, 8 s, 9 s, 10 s, 11 s, 12 s, 13 s. , 14s, 15s, 16s, 17s, 18s, 19s, 20s, 21s, 22s, 23s, 24s, 25s, 26s, 27s, 28s, 29s, 30s, 31s, 32s, 33s, 34s, 35s, 36s, 37s, 38s , 39s, 40s, 41s, 42s, 43s, 44s, 45s, 46s, 47s, 48s, 49s, 50s, 51s, 52s, 53s, 54s, 55s, 56s, 57s, 58s, 59s, or 60s. The sampling interval is a time interval with a time interval accuracy greater than 0.1 s, or even a higher time interval accuracy, such as 1.1 s, 1.1. 1s, etc. It should be noted that the above sampling interval is only a time interval for exemplifying the sampling of the energy consumption data, instead of indicating that the sampling interval uses only the above examples to perform the sampling processing of the energy consumption data. With the hardware improvement of the electric energy meter and/or the reduction of the data transmission cost, the sampling interval can be reduced to within 1 s, such as 0.5 s, etc., and the non-invasive energy consumption detecting system 5 can be provided in more detail according to the sampling device. The energy consumption sequence performs more accurate energy consumption detection, in other words, those skilled in the art, in the context of knowing the innovative spirit of the present application, should adjust the above-mentioned time interval according to the actual power meter performance and the application of the data transmission technology. Belongs to the scope covered by this application.

In order to collect the foregoing energy consumption data and form an energy consumption sequence for the change point detection module 51 to perform the change point detection, the energy consumption collection module 53 provides the energy consumption sequence obtained by sampling from the total power supply loop to the change point detection module 51. . The energy consumption collection module 53 is connected to the collection device data, and the collected energy consumption data is saved through a database. The database used by the energy consumption collection module 53 includes but is not limited to: oracle, SQL, and the like. The change point detection module 51 can perform batch extraction of the stored energy consumption data by the energy consumption collection module 53 to obtain an energy consumption sequence.

The non-intrusive energy consumption detecting system 5 performs energy consumption analysis on the acquired energy consumption sequence by performing the change point detecting module 51 and the energy consumption detecting module 52, and obtains a mapping of the operating state change event and the energy consumption change of the monitored device. Relationship, and performing energy consumption monitoring on the monitored device by using the mapping relationship. The energy consumption monitoring includes, but is not limited to, at least one of: detecting faults of the monitored equipment, measuring power consumption of the monitored equipment, and optimizing operation of the monitored equipment to reduce energy consumption.

Here, the monitored device is usually a device that is of high concern for the site management or the property owner, such as a frequency conversion device, a large power consumption device, etc., but is not limited thereto, and can be designed according to actual needs by those skilled in the art. The energy consumption detection scheme described in the present application performs energy consumption detection on the operating state events of the fixed frequency device and the small power consumption device. Here, the high power consumption device is exemplified by a rated power, or a device with a maximum power of more than kilowatts, or a device with a rated power and a maximum power of more than kilowatts. The low power consumption device is exemplified by a rated power, or a device having a maximum power of less than kilowatts, or a device having a rated power and a maximum power of less than kilowatts.

It can be seen that a change point may be caused by a change in the operating state of one device or a combination of operating states of multiple devices. We refer to a combination of a running state change or a change in the operating state of multiple devices that can cause a change point as a device operating state change event. In order to more accurately monitor the energy consumption of the devices of interest, we are more concerned with the device operating state change events caused by the changes in the operating state of these devices. Wherein, when the monitored device includes the variable frequency device, the device operating state change event includes a change in the operating state of the variable frequency device.

The non-intrusive energy consumption detecting system 5 starts the change point detecting module 51 after obtaining the energy consumption sequence.

The change point detecting module 51 is configured to perform a change point detection on the energy consumption sequence of the acquired load to obtain a change point sequence. The load includes all the loads on the loop where the energy consumption sequence is collected, including but not limited to: small appliances such as lamps, computers, printers, household appliances such as refrigerators and televisions, and monitored inverters such as elevators and central air conditioners. Or fixed frequency equipment, etc.

Here, the change point detecting module 51 can select a corresponding change point detecting mode according to the data type of the energy consumption sequence. For example, if the energy consumption sequence is a voltage data type, the change point detection may be performed by using a harmonic analysis method or the like. If the energy consumption sequence is a power data type, the change point detection may be performed by using Allen variance and load detection. In some embodiments, the change point detection module 51 performs a change point detection on each energy consumption data in the energy consumption sequence by using a Bayesian algorithm. For example, detecting the first energy consumption data with the initial condition and the posterior condition is the possibility of sampling before the change point. When the probability meets the sampling probability threshold before the change point, the first energy consumption of the sampling is determined. The time corresponding to the data is before the change point; then, according to the determined result of the first energy consumption data corresponding to the change point, the a priori condition and the posterior condition are adjusted and the second energy consumption data change point is detected The possibility of energy consumption data, when the possibility meets the pre-change condition, it is determined that the time corresponding to the second energy consumption data is still sampled before the change point. The energy consumption data is iteratively detected until it is detected that the sampling time corresponding to an energy consumption data is after the change point, and then the change point is determined to occur within the sampling interval of the energy consumption data before and after the change point. The Bayesian algorithm is used to detect a change point in the energy consumption sequence.

The energy consumption detecting module 52 is configured to map and identify a sequence of device operating state change events that cause the sequence of changes according to the energy consumption change information corresponding to the device operating state change obtained in advance to correspond to the load. The device performs energy consumption monitoring.

Since the actual energy consumption sequence can provide more realistic and complex energy consumption noise, and the above-mentioned energy consumption change information is determined to be idealized, the accuracy of establishing the above mapping relationship needs to be improved. To this end, in some embodiments, the energy consumption change information is obtained by machine learning. Here, since the present application uses non-intrusive energy consumption detection, the system described in the present application further includes a learning module that is executed before the energy consumption detecting module 52 or executed in parallel with the energy consumption detecting module 52 (not To the illustration).

The learning module is configured to perform machine learning by using the change point sequence to obtain the energy consumption change information. In some embodiments, the energy subsequence corresponding to each change point is obtained by accumulating for a period of time; according to the power consumption parameter of the monitored device, the electrical characteristics of the main electrical device in the device, etc., characteristic analysis of each energy subsequence is performed to obtain a device. The change in energy consumption corresponding to the change in operating status. In other embodiments, the energy subsequence corresponding to each change point is obtained by accumulating for a period of time; each energy subsequence is clustered, and then the energy consumption subsequence of the same classification is analyzed to obtain energy consumption change information; The power consumption parameter of the monitoring device and the electrical characteristics of the main electrical device in the device are matched, and the change of the operating state of the device is matched with the energy consumption change information of each category, so that the energy consumption change information corresponding to the change of the operating state of the device is obtained.

It should be noted that the learning module can perform continuous machine learning according to the obtained sequence of change points, and continuously update the obtained energy consumption change information, so that the energy consumption detecting module 52 can provide more accurate energy consumption change information. To improve the detection accuracy of the non-invasive energy consumption detection system 5.

After the energy consumption change information corresponding to the change of the operating state of the device is obtained, the energy consumption detecting module 52 performs the sequence of the device operating state change event that causes the change point sequence based on the energy consumption change information corresponding to the device operating state change. Mapping recognition.

Here, according to the monitored device and the number of changes in its operating state, a corresponding number of device operating state change events are preset; the energy subsequence corresponding to each change point in the change point sequence, energy consumption data or energy consumption The data change value is matched with each energy consumption change information, and each matching degree reflects the possibility P1 of each operation state change of each device at the change point time; and the device operation state change event generated at a single change point is determined based on each possibility P1. Possibility P2, and the possibility P3 associated with the combination of device operating state change events of a plurality of change points in the sequence of change points; the probability P2 corresponding to each device operating state change event based on the change point sequence And P3, obtaining a mapping relationship between each change point in the change point sequence and each device running state change event. Therefore, the energy consumption detecting module 52 can detect the energy consumption change when the operating state of the corresponding device changes according to the obtained mapping relationship.

Among them, different state transitions of the same device may be mutually exclusive. For example, a state transition from start to stop and a state transition from stop to start are not possible at the same time. In addition, during the operation of multiple devices based on timing generation, different state transitions of the same device and state transitions between different devices are mutually exclusive. For example, if the change point A1 is caused by the device D1 transitioning from the start to the stop state, the internal change point of this interval may not be initiated by the device D1 until the change point of the device D1 transition from the stop to the start state is not detected. Any state transition between a stop state transition, or device D1 operation or standby. Therefore, when each of the possibilities P1, P2, and P3 is calculated for each possible initial assignment, or when the possibilities P1, P2, and P3 are calculated, or when the mapping relationship is determined based on the possibilities P2 and P3, The adjustment of the corresponding possibility or the mapping relationship is performed based on the above exclusion condition, or the corresponding possibility and the mapping relationship are adjusted based on the above exclusion condition to improve the detection accuracy.

In some embodiments, based on the energy consumption change information and the energy consumption deviation of the energy consumption subsequence in the change point sequence, each change point in the change point sequence is mapped and identified with each device operation state change event. Specifically, the energy consumption detecting module 52 performs the following steps:

In an implementation manner, the energy consumption detecting module 52 further performs step S124: adjusting a change point and a device in the change point sequence based on at least one of a device energy consumption parameter obtained in advance and an influencing factor of device operation. The mapping relationship of running status change events. The energy consumption parameters of the device are exemplified by the rated power, the maximum power, and the like of the device, and other parameters related to the energy consumption of the device. The energy consumption parameter of the device may be obtained from a product specification, a specification, a user manual of the device, or may be provided by a park management party or a property management party. The influencing factors include, but are not limited to, weather, occupancy rate, equipment start-stop schedule, etc. These influencing factors may directly or indirectly become factors in the change of the operating state of the equipment. For example, if the weather is lower than t1, the maintenance personnel of the equipment turn on the heating equipment. The weather maintenance equipment only installs the fresh air equipment between t1 and t2 degrees. The equipment maintenance personnel turn on the cold water pump and the fan equipment when the weather is above t2 degrees. For another example, the equipment maintenance personnel control the start and stop of each equipment of the building air conditioning system according to the equipment start-stop schedule. The influencing factors can be obtained through a network from a third party service device, through a manually imported configuration file, or through a human machine interface.

In order to make the energy consumption detecting module 52 more suitable for a place, a specific example is that the non-invasive energy consumption detecting system 5 further includes an input module (not shown). The input module is configured to pre-provide an input interface of at least one of the device energy consumption parameter and the influencing factor of the device operation to obtain a corresponding device energy consumption parameter and an influencing factor of the device operation. For example, the input module includes an input interface that can input the energy consumption parameters of the monitored devices and the influencing factors of the device operation in the monitoring platform interface provided to the user. When the user clicks the submit button, the energy consumption detecting module 52 Obtain the device energy consumption parameters in the interface and the influencing factors of the device operation. The energy consumption detecting module 52 assigns the acquired device energy consumption parameter and the influencing factor of the device operation to corresponding parameters in the preset statistical algorithm.

Since the energy consumption parameters of the equipment and the influencing factors of the equipment operation may affect the operation state of the equipment, it may optimize multiple steps of the mapping relationship determination process. For example, according to the weather information in the obtained influencing factors, the weather is less than 10 degrees, and the energy consumption detecting module 52 may only calculate the energy deviation of the energy subsequence corresponding to each change point and the operating state change of the heating device. For another example, according to the acquired influencing factors, the equipment running schedule is to start the cold water pump at 8:00 am on Monday, and the energy consumption detecting module 52 can increase the change point generated during the 7:50-8:10 period and include the cold water pump. The possibility of stopping the mapping of events to the start state transition, or increasing the transition of the cold water pump from stop to start state in each event, or including the transition of the cold water pump from stop to start state, or both events of the above several cases Evaluate the weight. For another example, according to the rated power in the energy consumption parameter of the device, the error between the obtained energy consumption deviation and the rated power is evaluated. If the error is determined to be within the preset reliability range, the corresponding change point may be improved. Possibilities and/or weights caused by changes in the operating state of the respective device.

It should be noted that, through the description of the above embodiments, those skilled in the art can clearly understand that some or all of the application can be implemented by software and in combination with a necessary general hardware platform. Based on such understanding, portions of the technical solution of the present application that contribute in essence or to the prior art may be embodied in the form of a software product, which may include one or more of the executable instructions for storing the machine thereon. A machine-readable medium that, when executed by one or more machines, such as a computer, computer network, or other electronic device, can cause the one or more machines to perform operations in accordance with embodiments of the present application. The machine-readable medium can include, but is not limited to, a floppy disk, an optical disk, a CD-ROM (Compact Disk-Read Only Memory), a magneto-optical disk, a ROM (Read Only Memory), a RAM (Random Access Memory), an EPROM (erasable) In addition to programmable read only memory, EEPROM (Electrically Erasable Programmable Read Only Memory), magnetic or optical cards, flash memory, or other types of media/machine readable media suitable for storing machine executable instructions.

This application can be used in a variety of general purpose or special purpose computing system environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor based systems, set-top boxes, programmable consumer electronics devices, network PCs, small computers, mainframe computers, including A distributed computing environment of any of the above systems or devices.

The application can be described in the general context of computer-executable instructions executed by a computer, such as a program module. Generally, program modules include routines, programs, objects, components, data structures, and the like that perform particular tasks or implement particular abstract data types. The present application can also be practiced in distributed computing environments where tasks are performed by remote processing devices that are connected through a communication network. In a distributed computing environment, program modules can be located in both local and remote computer storage media including storage devices.

It should be noted that those skilled in the art may understand that some of the above components may be programmable logic devices, including: Programmable Array Logic (PAL), Generic Array Logic (GAL), and on-site One or more of a Field-Programmable Gate Array (FPGA) and a Complex Programmable Logic Device (CPLD), the present invention/invention does not specifically limit this.

The present application has been disclosed in the above preferred embodiments, but it is not intended to limit the present application. Any person skilled in the art can use the methods and technical contents disclosed above to apply to the present application without departing from the spirit and scope of the present application. The technical solutions make possible changes and modifications. Therefore, any simple modifications, equivalent changes, and modifications made to the above embodiments in accordance with the technical spirit of the present application are not included in the technical solutions of the present application. protected range.

Claims

A non-invasive energy consumption detecting method, characterized in that the method comprises the following steps:

Performing a change point detection on the energy consumption sequence of the acquired load to obtain a sequence of change points;

And mapping, according to the energy consumption change information corresponding to the change of the operating state of the device, the device operating state change event sequence that causes the change sequence to perform energy consumption monitoring on the corresponding device in the load.
The non-invasive energy consumption detecting method according to claim 1, further comprising the step of acquiring the energy consumption sequence by a collecting device connected to the main power supply circuit.
The non-invasive energy consumption detecting method according to claim 1, wherein the sampling interval of the adjacent energy consumption data in the energy consumption sequence is in the range of (0, 60) s.
The non-invasive energy consumption detecting method according to claim 1, wherein the manner of performing change point detection on the energy consumption sequence of the acquired load comprises: using a Bayesian algorithm for each of the energy consumption sequences The energy consumption data is subjected to change point detection.
The method for detecting non-invasive energy consumption according to claim 1, wherein the sequence of change points is obtained by the plurality of energy subsequences including the detected change points acquired in the energy consumption sequence. The sequence of the composition.
The non-invasive energy consumption detecting method according to claim 1 or 5, wherein the device operating state that causes the change point sequence is based on the energy consumption change information corresponding to the pre-determined device operating state change The manner in which the change event sequence is mapped and identified includes: performing, according to the energy consumption change information, an energy consumption deviation of the energy consumption subsequence in the change point sequence, performing an event change event of each change point and each device in the change point sequence Mapping recognition.
The non-invasive energy consumption detecting method according to claim 1, wherein the device operating state change event that causes the change point sequence is based on the energy consumption change information corresponding to the pre-determined device operating state change. The method for performing sequence mapping identification includes: adjusting a mapping relationship between the change point of the change point sequence and a device running state change event based on at least one of a pre-obtained device energy consumption parameter and an influencing factor of the device operation.
The non-invasive energy consumption detecting method according to claim 7, further comprising an input interface that provides at least one of the device energy consumption parameter and the influencing factor of the device operation in advance to obtain the corresponding device energy consumption. Steps for influencing factors on parameters and device operation.
The non-invasive energy consumption detecting method according to claim 1, wherein the device operating state change comprises at least one of a device start-stop state transition and a transition between the plurality of operating states during operation of the device.
The non-intrusive energy consumption detecting method according to claim 1, wherein the energy consumption change information corresponding to the device operating state change is obtained in a non-intrusive manner.
The non-invasive energy consumption detecting method according to claim 10, further comprising the step of performing machine learning on the change point sequence to obtain the energy consumption change information.
The non-invasive energy consumption detecting method according to claim 1, wherein the device operating state change event comprises an operating state change of the frequency conversion device.
A non-invasive energy consumption detection system, comprising:

a collecting device for collecting energy consumption data of the load;

At least one storage device for storing energy consumption data collected by the collection device and at least one computer program;

At least one processing device for executing the at least one computer program such that the energy consumption detection of the method of any of claims 1-12 is performed with an energy consumption sequence formed by the energy consumption data.
The non-invasive energy consumption detecting system according to claim 13, wherein the collecting device collects energy consumption data within an interval of (0, 60) s.
The non-invasive energy consumption detecting system according to claim 13, wherein the collecting device is connected to an electric energy meter connected to the main power supply circuit.
The non-invasive energy consumption detecting system according to claim 13, wherein the energy consumption data collected by the collecting device comprises energy consumption data generated by the operation of the frequency converting device.
A server is configured to be in communication with a collection device, wherein the collection device is configured to collect energy consumption data of a load on a total power supply loop, and the server includes:

One or more processors;

One or more memories for storing energy consumption data provided by the collection device and at least one computer program;

When the one or more computer programs are executed by the one or more processors, causing the one or more processors to perform the energy consumption sequence formed by the energy consumption data as in claims 1-12 Energy consumption detection by any of the methods described.
A non-invasive energy consumption detection system, comprising:

a change point detecting module, configured to perform a change point detection on the energy consumption sequence of the acquired load to obtain a change point sequence;

The energy consumption detecting module is configured to perform mapping mapping on the device operating state change event sequence that causes the change point sequence to perform the corresponding device in the load based on the energy consumption change information corresponding to the device operating state change obtained in advance Energy consumption monitoring.
The non-invasive energy consumption detecting system according to claim 18, further comprising an energy consumption collecting module, configured to obtain an energy consumption sequence sampled from the total power supply loop.
The non-invasive energy consumption detecting system according to claim 18, wherein the sampling interval of the adjacent energy consumption data in the energy consumption sequence is in the range of (0, 60) s.
The non-intrusive energy consumption detecting system according to claim 18, wherein the change point detecting module is configured to perform a change point detection on each energy consumption data in the energy consumption sequence by using a Bayesian algorithm.
The non-invasive energy consumption detecting system according to claim 18, wherein the sequence of change points is obtained by the plurality of energy subsequences including the detected change points acquired in the energy consumption sequence The sequence of the composition.
The non-invasive energy consumption detecting system according to claim 18 or 22, wherein the energy consumption detecting module is configured to perform energy consumption deviation based on the energy consumption change information and the energy consumption subsequence in the change point sequence And mapping and identifying each change point in the change point sequence and each device running state change event.
The non-invasive energy consumption detecting system according to claim 18, wherein the energy consumption detecting module is configured to adjust the at least one of a device energy consumption parameter obtained in advance and an impact factor of device operation The mapping relationship between the change point in the change point sequence and the device running state change event.
The non-invasive energy consumption detecting system according to claim 24, further comprising an input module, configured to pre-provide an input interface of at least one of the device energy consumption parameter and the influencing factor of the device operation to advance Obtaining at least one of the energy consumption parameter of the device and the influencing factor of the operation of the device.
The non-invasive energy consumption detecting system according to claim 18, wherein the device operating state change comprises at least one of a device start-stop state transition and a transition between the plurality of operating states during operation of the device.
The non-intrusive energy consumption detecting system according to claim 18, wherein the energy consumption change information corresponding to the device operating state change is obtained in a non-intrusive manner.
The non-invasive energy consumption detecting system according to claim 27, further comprising a learning module, configured to perform machine learning by using the change point sequence to obtain the energy consumption change information.
The non-invasive energy consumption detecting system according to claim 18, wherein the device operating state change event comprises an operating state change of the variable frequency device.