WO2020063686A1 - 一种热负荷预测方法、装置、可读介质及电子设备 - Google Patents

一种热负荷预测方法、装置、可读介质及电子设备 Download PDF

Info

Publication number
WO2020063686A1
WO2020063686A1 PCT/CN2019/107928 CN2019107928W WO2020063686A1 WO 2020063686 A1 WO2020063686 A1 WO 2020063686A1 CN 2019107928 W CN2019107928 W CN 2019107928W WO 2020063686 A1 WO2020063686 A1 WO 2020063686A1
Authority
WO
WIPO (PCT)
Prior art keywords
state
prediction
time period
steam
future time
Prior art date
Application number
PCT/CN2019/107928
Other languages
English (en)
French (fr)
Inventor
刘胜伟
黄信
Original Assignee
新智数字科技有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 新智数字科技有限公司 filed Critical 新智数字科技有限公司
Priority to EP19865708.2A priority Critical patent/EP3822839A4/en
Priority to US17/058,137 priority patent/US11887020B2/en
Priority to SG11202102668YA priority patent/SG11202102668YA/en
Priority to JP2021514404A priority patent/JP7426992B2/ja
Publication of WO2020063686A1 publication Critical patent/WO2020063686A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N7/00—Computing arrangements based on specific mathematical models
    • G06N7/01—Probabilistic graphical models, e.g. probabilistic networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2111/00—Details relating to CAD techniques
    • G06F2111/08—Probabilistic or stochastic CAD
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
    • G06F2119/08—Thermal analysis or thermal optimisation

Definitions

  • the invention relates to the technical field of electrical engineering, and in particular, to a method, a device, a readable medium, and an electronic device for predicting a thermal load.
  • an autoregressive integrated moving average model (ARIMA model) is usually used to predict the respective thermal load magnitudes of steam users when using steam from steam equipment in various future time periods.
  • the ARIMA model can only predict the magnitude of the heat load when steam users use steam in various future time periods, that is, a specific value is predicted. For irregular use of steam equipment in multiple consecutive historical time periods, there may be a large difference between one or more of the predicted heat load levels and the current heat load level corresponding to the steam user's actual use of steam equipment steam in the corresponding future time period. The difference is that the single point error is too large, which makes it impossible to efficiently schedule the steam of the steam equipment in the subsequent process.
  • the invention provides a heat load prediction method, device, readable medium and electronic equipment, which can realize the prediction of the heat load interval corresponding to the steam user when using the steam of the steam equipment in each future time period, which is convenient for subsequent processes. Steam equipment steam for more efficient scheduling.
  • the present invention provides a thermal load prediction method, including:
  • the state confidence probability of the future time period with respect to the various predicted states, and the various predicted states determines a heat load interval corresponding to the target steam user when using the steam equipment steam in the future time period.
  • the determining the relative prediction error corresponding to each of the test time periods according to the first heat load magnitude corresponding to each of the test time periods includes:
  • E i represents the relative prediction error corresponding to the i-th test period
  • F i represents the actual heat load magnitude corresponding to the i-th test period
  • f i represents the first corresponding to the i-th test period.
  • the state confidence probability of the predicted state includes:
  • the confidence probability of each future time period with respect to the various predicted states is determined.
  • the determining, according to the initial probability matrix and the state transition probability matrix, the confidence probability of each future time period relative to the various predicted states includes:
  • X (n) represents the confidence probability matrix of the n-th future time period relative to the various predicted states with the cut-off point of the current time period as the starting time point
  • X (0) represents the corresponding The initial probability matrix and P characterize the state transition probability matrix
  • the confidence probability of the future time period relative to the various predicted states is extracted from a confidence probability matrix corresponding to the future time period.
  • the heat load interval corresponding to the target steam user using the steam equipment steam in the future time period includes:
  • Determining, according to the error threshold corresponding to the target predicted state and the second heat load level corresponding to the future time period, the heat corresponding to the target steam user when using steam equipment steam in the future time period Load interval including:
  • y represents the upper or lower critical value corresponding to the future time period
  • the second heat load magnitude corresponding to the future time period described in the Y table
  • e represents the error threshold value corresponding to the target prediction state.
  • a heat load interval corresponding to the target steam user when using the steam equipment steam in the future time period is determined.
  • the multiple prediction states specifically include: extremely overestimated state, overestimated state, normal state, underestimated state, and extremely underestimated state; among them,
  • the error threshold corresponding to the extremely overestimated state is specifically greater than 10%
  • the error threshold corresponding to the overestimation state is specifically greater than 5% and not greater than 10%;
  • the error threshold corresponding to the normal state is specifically not less than -5% and not more than 5%;
  • the error threshold corresponding to the underestimated state is specifically not less than -10% and less than -5%;
  • the error threshold corresponding to the extremely underestimated state is specifically less than -10%.
  • the present invention provides a thermal load prediction device, including:
  • a pre-processing module for setting various prediction states and their corresponding error thresholds to form a prediction model corresponding to the target steam user
  • a model calling module is used to call the prediction model to predict the first heat load magnitude corresponding to the target steam user when using the steam equipment steam in multiple test periods; call the prediction model to predict the target user Second heat load levels corresponding to the steam equipment steam used in each of said future time periods;
  • An error processing module configured to determine a relative prediction error corresponding to each of the test time periods according to a first heat load magnitude corresponding to each of the test time periods;
  • the state probability determination module forms a state transition probability matrix according to a relative prediction error corresponding to each of the test time periods and an error threshold corresponding to each of the predicted states, and determines each future time period according to the state transition probability matrix. State confidence probability with respect to various said predicted states;
  • a thermal load interval prediction module is configured to, for each of the future time periods, according to a second thermal load magnitude corresponding to the future time period, and state confidence of the future time period relative to various predicted states.
  • the probability and the error thresholds corresponding to the various predicted states respectively determine the heat load interval corresponding to the target steam user when using the steam equipment steam in the future time period.
  • the present invention provides a readable medium including an execution instruction.
  • the processor of the electronic device executes the execution instruction, the electronic device executes the method according to any one of the first aspects.
  • the present invention provides an electronic device, including: a processor, a memory, and a bus; the memory is used to store execution instructions, and the processor and the memory are connected through the bus.
  • the processor executes the execution instructions stored in the memory, so that the processor executes the method according to any one of the first aspects.
  • the invention provides a method, a device, a readable medium, and an electronic device for predicting a thermal load.
  • the method sets multiple prediction states and their corresponding error thresholds, and forms a prediction model corresponding to a target steam user. After the model predicts the first heat load level corresponding to the target steam user when using the steam of the steam equipment in multiple test time periods, each test time period can be determined according to the first heat load level corresponding to each test time period.
  • the corresponding corresponding relative prediction error, and then the state transition probability matrix is formed according to the relative prediction error corresponding to each test period, and the state probability of each future time period relative to various predicted states is determined according to the state transition probability matrix, and called again
  • the prediction model predicts the second heat load level corresponding to the target user when using the steam of the steam equipment in each future time period, it can be targeted for each future time period according to its corresponding second heat load level, its State probability relative to various prediction states and various predictions
  • the error thresholds corresponding to the respective states determine the heat load interval corresponding to the target steam user when using steam from the steam equipment in this future period of time, to optimize the predicted second heat load magnitude to avoid single-point errors. Large phenomena can facilitate more efficient scheduling of steam equipment steam in subsequent processes.
  • FIG. 1 is a schematic flowchart of a thermal load prediction method according to an embodiment of the present invention
  • FIG. 2 is a schematic flowchart of another thermal load prediction method according to an embodiment of the present invention.
  • FIG. 3 is a schematic structural diagram of a thermal load prediction device according to an embodiment of the present invention.
  • FIG. 4 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.
  • an embodiment of the present invention provides a thermal load prediction method, including:
  • Step 101 Set various prediction states and their corresponding error thresholds to form a prediction model corresponding to the target steam user;
  • step 102 the prediction model is called to predict the first heat load magnitudes corresponding to the target steam user when using the steam equipment steam in a plurality of test periods;
  • Step 103 Determine a relative prediction error corresponding to each of the test time periods according to a first heat load magnitude corresponding to each of the test time periods;
  • Step 104 Form a state transition probability matrix according to the relative prediction error corresponding to each of the test time periods and the error threshold corresponding to each of the predicted states, and determine the relative of each future time period with respect to the state transition probability matrix. State confidence probability of various said predicted states;
  • step 105 the prediction model is called to predict the second heat load level corresponding to the target user when using the steam equipment steam in each of the future time periods;
  • Step 106 For each of the future time periods, according to the second heat load magnitude corresponding to the future time period, the state confidence probability of the future time period with respect to the various predicted states, and various The error thresholds corresponding to the predicted states respectively determine a heat load interval corresponding to the target steam user when using the steam equipment steam in the future time period.
  • the method sets a plurality of prediction states and their corresponding error thresholds, and forms a prediction model corresponding to the target steam user, and calls the prediction model to predict the target steam user at multiple test time periods.
  • the relative prediction error corresponding to each test time period can be determined according to the first heat load magnitude corresponding to each test time period, and then according to each
  • the relative prediction errors corresponding to the test time periods form a state transition probability matrix, and the state probability of each future time period relative to various prediction states is determined according to the state transition probability matrix.
  • the prediction model is called again to predict the target user in each future time period.
  • the second heat load magnitude corresponding to the steam equipment steam After using the second heat load magnitude corresponding to the steam equipment steam, it can be targeted for each future time period according to its corresponding second heat load magnitude, its state probability relative to various predicted states, and various The error threshold corresponding to each predicted state determines the target steam The heat load corresponding to the time when the user uses the steam equipment steam in this future period of time, to optimize the predicted second heat load magnitude to avoid excessive single point error, which can facilitate the steam equipment steam in the subsequent process. For more efficient scheduling.
  • the prediction model can be an auto-regressive integrated moving average model (ARIMA model) or a GBRT (Gradient Boost Boost Regression Tree) prediction model.
  • ARIMA model auto-regressive integrated moving average model
  • GBRT GBRT
  • Gradient Boost Boost Regression Tree Boost Regression Tree
  • time lengths of the "historical time period”, “short test time”, and “future time period” described in any of the embodiments of the present invention are the same, and the specific length may be 1 hour; obviously In some special business scenarios, the length of the time period can also be set to other values.
  • the multiple prediction states specifically include: extremely overestimated state, overestimated state, normal state, underestimated state, and extremely underestimated state; wherein,
  • the error threshold corresponding to the extremely overestimated state is specifically greater than 10%
  • the error threshold corresponding to the overestimation state is specifically greater than 5% and not greater than 10%;
  • the error threshold corresponding to the normal state is specifically not less than -5% and not more than 5%;
  • the error threshold corresponding to the underestimated state is specifically not less than -10% and less than -5%;
  • the error threshold corresponding to the extremely underestimated state is specifically less than -10%.
  • the multiple prediction states set in this embodiment can be increased or decreased by the user in accordance with the different needs of the actual business scenario.
  • Each prediction state is configured with a different error threshold, which is verified after a limited number of implementations.
  • the corresponding error for each prediction state When the threshold value is the threshold value described in the foregoing embodiment, it can be ensured that the actual heat load corresponding to the target steam user when using the steam of the steam equipment in each future time period is located in the predicted value through the technical solution provided by the embodiment of the present invention. Within the corresponding heat load interval, that is, to ensure that the technical solution provided by the embodiment of the present invention can more accurately predict the respective heat load levels corresponding to the target steam user when using the steam equipment steam in each future time period.
  • S1 is used to characterize the overestimation state
  • S2 is used to characterize the overestimation state
  • S3 is used to characterize the normal state
  • S4 is used to characterize the underestimation state
  • S5 is used to characterize the extremely underestimation state.
  • the determining the relative prediction error corresponding to each of the test time periods according to the first heat load magnitude corresponding to each of the test time periods includes:
  • E i represents the relative prediction error corresponding to the i-th test period
  • F i represents the actual heat load magnitude corresponding to the i-th test period
  • f i represents the first corresponding to the i-th test period.
  • the first heat load level corresponding to the target steam user when using the steam equipment steam in each test time period is predicted by the prediction model, and the target steam user is used to use the steam equipment steam in each test time period.
  • the first heat load magnitude and the actual heat load magnitude corresponding to each test period are further processed by Formula 1, and each relative prediction error obtained can be used
  • the prediction state between the predicted first heat load magnitude and the actual heat load magnitude corresponding to the corresponding test period is evaluated, and the error magnitude of the first heat load magnitude corresponding to the corresponding test period is measured.
  • the state transition probability matrix is formed according to the relative prediction error corresponding to each of the test time periods and the error threshold corresponding to each of the predicted states, and the state transition probability is based on the state transition probability.
  • the matrix determines the state confidence probability of each future time period relative to the various predicted states, including:
  • the confidence probability of each future time period with respect to the various predicted states is determined.
  • the corresponding prediction states S1, S2, S3, S4, and S5 corresponding to the foregoing embodiments correspond to
  • the error threshold can be determined to be 6% within the error threshold corresponding to the overestimated state S2.
  • the overestimated state S2 can be determined as the trusted prediction state corresponding to the i-th test period.
  • the prediction state transition type specifically refers to a situation in which one prediction state transitions to the current prediction state or other prediction states among the multiple prediction states that are set; specifically, the five prediction states described in the above embodiment S1, S2, S3, S4, and S5 are examples.
  • S2, S3, S4, and S5 also correspond to five types of transitions, each of which is a type of predicted state transition. Then, there are 25 types of predicted state transitions.
  • the trusted prediction state corresponding to the i-1th test period is S1
  • the trusted prediction state corresponding to the ith test period is S2
  • the number of state transitions corresponding to the predicted state transition type from S1 to S2 can be increased by 1; after performing the foregoing processing for every two adjacent test time periods, the corresponding prediction state transition types can be obtained respectively.
  • the ratio of the corresponding number of state transitions to the total number of test periods n is the state transition probability P corresponding to the state transition type.
  • the position of the state transition probability corresponding to each state transition type in the formed state transition probability matrix can be referred to the following Table 1:
  • state transition probability P S1-S2 in Table 1 includes the state transition probability corresponding to the characteristic state transition type “S1 to S2 transition”. Those skilled in the art can understand each of the records in Table 1 according to similar rules. State transition probability.
  • each state transition probability in Table 1 that is, when the state transition probability is used as an element in the state transition probability matrix, the element corresponds to the row and column position in the state transition probability.
  • the initial probability matrix has 5 elements, 5 The elements are the initial probabilities of the prediction states S1, S2, S3, S4, and S5 in order, and only the initial probabilities of the trusted prediction states corresponding to the current time period are 1, and the initial probabilities of the other prediction states are all 0.
  • the confidence probability of each future time period relative to various prediction states is determined according to the initial probability matrix and the state transition probability matrix, for a future time period.
  • the probability means that when the prediction model formed is used to predict the second heat load level corresponding to the target steam user who has used steam equipment steam in the future time period, the predicted second heat load level is The higher the probability that the corresponding prediction state is the current prediction state, it is convenient for subsequent processes to use the confidence probability of the future time period relative to various prediction states to predict the second corresponding to the future time period predicted by the prediction model.
  • the magnitude of the heat load is optimized to more accurately determine the heat load interval corresponding to the target steam user when using the steam of the steam equipment in the future time period.
  • determining the confidence probability of each future time period with respect to various prediction states according to the initial probability matrix and the state transition probability matrix includes:
  • X (n) represents the confidence probability matrix of the n-th future time period relative to the various predicted states with the cut-off point of the current time period as the starting time point
  • X (0) represents the corresponding The initial probability matrix and P characterize the state transition probability matrix
  • the confidence probability of the future time period relative to various prediction states is extracted from a confidence probability matrix corresponding to the future time period.
  • the multiple prediction states described in the foregoing embodiment specifically include S1, S2, S3, S4, and S5 as an example. It may include 5 elements, and the 5 elements are the confidence probability of the nth future time period relative to the prediction states S1, S2, S3, S4, and S5 in order.
  • the error thresholds corresponding to the states to determine the heat load interval corresponding to the target steam user when using the steam equipment steam in the future time period include:
  • the determining the target steam user at the future time according to the error threshold corresponding to the target prediction state and the second heat load level corresponding to the future time period includes:
  • y represents the upper or lower critical value corresponding to the future time period
  • the second heat load magnitude corresponding to the future time period described in the Y table
  • e represents the error threshold value corresponding to the target prediction state.
  • a heat load interval corresponding to the target steam user when using the steam equipment steam in the future time period is determined.
  • the confidence probability of a future time period relative to a certain current prediction state is higher, it means that the second time corresponding to the target steam user using the steam equipment steam in the future time period when the prediction model is called
  • the more likely the predicted state corresponding to the predicted second heat load magnitude is the current predicted state; therefore, after the predicted state corresponding to the maximum confidence probability is determined as the target state, The error threshold corresponding to the target prediction state.
  • the prediction model When the user predicts the second heat load level corresponding to the time when the user uses the steam of the steam equipment in the jth future time period, the predicted second heat load level corresponds to a higher possibility of overestimation; specifically, see the foregoing
  • the error threshold corresponding to the predicted state S2 is greater than 5% and not greater than 10%, that is, the upper extreme value (the maximum value or the limit value approached by the maximum value) corresponding to the overestimated state S2 is 10%, and the lower The limit value (the limit value approached by the minimum value or the minimum value) is 5%, and the second heat load magnitude corresponding to the j-th future time period predicted by the prediction model is Y (the predicted Y value is overestimated Probability is extremely high), then, through the above formula 3,
  • the target steam user when scheduling for steam in the subsequent process, if the second steam load level Y predicted by the prediction model is used to schedule the steam for the target steam user in the ith future time period, it may result in the target steam user
  • the minimum thermal load demand for the target steam user may be 0.91Y
  • the highest The heat load demand may be 0.95Y
  • the median heat load demand is 0.9Y.
  • the target steam user may be targeted at the minimum heat load demand, maximum heat load demand, or median heat load demand of the target steam user in combination with the actual business scenario.
  • an embodiment of the present invention provides another method for predicting a heat load, which specifically includes the following steps.
  • Step 201 Set a plurality of prediction states and their corresponding error thresholds, and form a prediction model according to the historical heat load magnitudes corresponding to the target user when using the steam equipment steam in a plurality of consecutive historical time periods.
  • step 202 a prediction model is called to predict the first heat load levels corresponding to the target steam user when using the steam equipment steam in a plurality of consecutive test periods.
  • Step 203 Obtain the actual heat load levels corresponding to the target steam user when using the steam of the steam equipment in each test period.
  • Step 204 Calculate a relative prediction error corresponding to each test period.
  • Step 204 may be specifically implemented in combination with Formula 1 provided in any embodiment of the present invention.
  • Step 205 Determine the credible prediction state corresponding to each test time period according to the relative prediction error corresponding to each test time period and the corresponding error threshold of each prediction state.
  • Step 206 Determine the number of state transitions corresponding to each prediction state transition type according to the trusted prediction states corresponding to each two adjacent test time periods.
  • Step 207 Determine the state transition probability corresponding to each state transition type according to the state transition times corresponding to each state transition type.
  • Step 208 Use a state transition probability corresponding to each state transition type to form a state transition probability matrix.
  • Step 209 Determine an initial probability matrix of the current time period relative to various prediction states.
  • Step 210 Calculate a confidence probability matrix for each future time period with respect to various prediction states.
  • Step 210 may be specifically implemented in combination with Formula 2 provided in any embodiment of the present invention.
  • Step 211 For each future time period, the confidence probability of the future time period relative to various prediction states is extracted from the confidence probability matrix corresponding to the future time period.
  • the subsequent steps only perform corresponding processing for one future time period to achieve the determination of the heat load interval corresponding to the target steam user when using the steam equipment steam in the future time period.
  • Step 212 Determine the maximum confidence probability from the confidence probability of the future time period relative to various prediction states, and determine the prediction state corresponding to the maximum confidence probability as the target prediction state.
  • Step 213 Determine the heat load interval corresponding to the target steam user when using the steam equipment steam in the future time period according to the error threshold corresponding to the target prediction state and the second heat load level corresponding to the future time period.
  • the heat load interval corresponding to the target steam user using the steam equipment steam in the future time period can be determined by specifically combining the formula 3 provided in any of the foregoing embodiments.
  • the embodiment of the present invention further provides a thermal load prediction device, including:
  • the pre-processing module 301 is configured to set various prediction states and their corresponding error thresholds to form a prediction model corresponding to the target steam user;
  • a model calling module 302 is configured to call the prediction model to predict the first heat load magnitude corresponding to the target steam user when using the steam equipment steam in a plurality of test periods; call the prediction model to predict the target The second heat load magnitudes corresponding to the user when using the steam of the steam equipment in each of the future time periods;
  • An error processing module 303 configured to determine a relative prediction error corresponding to each of the test time periods according to a first heat load magnitude corresponding to each of the test time periods;
  • the state probability determination module 304 forms a state transition probability matrix according to the relative prediction error corresponding to each of the test time periods and an error threshold corresponding to each of the predicted states, and determines each future time according to the state transition probability matrix.
  • the thermal load interval prediction module 305 is configured to, for each of the future time periods, according to a second thermal load magnitude corresponding to the future time period, and a state of the future time period relative to various predicted states.
  • the confidence probability and the error thresholds corresponding to the various predicted states respectively determine the heat load interval corresponding to the target steam user when using the steam equipment steam in the future time period.
  • FIG. 4 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.
  • the electronic device includes a processor, and optionally also includes an internal bus, a network interface, and a memory.
  • the memory may include a memory, such as a high-speed random access memory (Random-Access Memory, RAM), and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.
  • RAM random access memory
  • non-volatile memory such as at least one disk memory.
  • the electronic device may also include hardware required for other services.
  • the processor, network interface and memory can be connected to each other through an internal bus, which can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus, or an EISA (Extended Industry Standard Architecture (Extended Industry Standard Architecture) bus and so on.
  • the bus can be divided into an address bus, a data bus, a control bus, and the like. For ease of representation, only a two-way arrow is used in FIG. 4, but it does not mean that there is only one bus or one type of bus.
  • the program may include program code, where the program code includes a computer operation instruction.
  • the memory may include memory and non-volatile memory, and provide instructions and data to the processor.
  • the processor reads the corresponding computer program from the non-volatile memory into the memory and then runs, and may also obtain the corresponding computer program from other devices to form a thermal load at the logical level. Forecasting device.
  • the processor executes a program stored in the memory to implement the thermal load prediction method provided in any embodiment of the present invention through the executed program.
  • the method executed by the thermal load prediction apparatus provided in the foregoing embodiment may be applied to a processor, or implemented by a processor.
  • the processor may be an integrated circuit chip with signal processing capabilities.
  • each step of the above method may be completed by an integrated logic circuit of hardware in a processor or an instruction in a form of software.
  • the aforementioned processor may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc .; it may also be a digital signal processor (DSP), special integration Circuit (Application Specific Integrated Circuit, ASIC), Field-Programmable Gate Array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.
  • DSP digital signal processor
  • ASIC Application Specific Integrated Circuit
  • FPGA Field-Programmable Gate Array
  • a general-purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
  • the steps of the method disclosed in combination with the embodiments of the present invention may be directly implemented by a hardware decoding processor, or may be performed by using a combination of hardware and software modules in the decoding processor.
  • the software module may be located in a mature storage medium such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, or an electrically erasable programmable memory, a register, and the like.
  • the storage medium is located in a memory, and the processor reads the information in the memory and completes the steps of the foregoing method in combination with its hardware.
  • An embodiment of the present invention further provides a computer-readable storage medium, where the computer-readable storage medium stores one or more programs, the one or more programs include instructions, and the instructions are executed by an electronic device including multiple application programs At this time, the electronic device can be caused to execute the thermal load prediction method provided in any embodiment of the present invention, and is specifically configured to execute the method shown in FIG. 1 and / or FIG. 2.
  • the system, device, module, or unit described in the foregoing embodiments may be specifically implemented by a computer chip or entity, or a product with a certain function.
  • a typical implementation device is a computer.
  • the computer may be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or A combination of any of these devices.
  • the embodiments of the present invention may be provided as a method, a system, or a computer program product. Therefore, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
  • computer-usable storage media including, but not limited to, disk storage, CD-ROM, optical storage, etc.
  • These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to work in a particular manner such that the instructions stored in the computer-readable memory produce a manufactured article including an instruction device, the instructions
  • the device implements the functions specified in one or more flowcharts and / or one or more blocks of the block diagram.
  • These computer program instructions can also be loaded on a computer or other programmable data processing device, so that a series of steps can be performed on the computer or other programmable device to produce a computer-implemented process, which can be executed on the computer or other programmable device.
  • the instructions provide steps for implementing the functions specified in one or more flowcharts and / or one or more blocks of the block diagrams.
  • a computing device includes one or more processors (CPUs), input / output interfaces, network interfaces, and memory.
  • processors CPUs
  • input / output interfaces output interfaces
  • network interfaces network interfaces
  • memory volatile and non-volatile memory
  • Memory may include non-persistent memory, random access memory (RAM), and / or non-volatile memory in computer-readable media, such as read-only memory (ROM) or flash memory (flash RAM). Memory is an example of a computer-readable medium.
  • RAM random access memory
  • ROM read-only memory
  • flash RAM flash memory
  • Computer-readable media includes both permanent and non-persistent, removable and non-removable media.
  • Information can be stored by any method or technology.
  • Information may be computer-readable instructions, data structures, modules of a program, or other data.
  • Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), and read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, read-only disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, Magnetic tape cartridges, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media may be used to store information that can be accessed by computing devices.
  • computer-readable media does not include temporary computer-readable media, such as modulated data signals and carrier waves.
  • the embodiments of the present invention may be provided as a method, a system, or a computer program product. Therefore, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
  • computer-usable storage media including, but not limited to, disk storage, CD-ROM, optical storage, etc.
  • the invention may be described in the general context of computer-executable instructions executed by a computer, such as program modules.
  • program modules include routines, programs, objects, components, data structures, and so on that perform specific tasks or implement specific abstract data types.
  • the invention can also be practiced in distributed computing environments in which tasks are performed by remote processing devices connected through a communication network.
  • program modules may be located in local and remote computer storage media, including storage devices.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Evolutionary Computation (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Geometry (AREA)
  • Computer Hardware Design (AREA)
  • Artificial Intelligence (AREA)
  • Software Systems (AREA)
  • Medical Informatics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Data Mining & Analysis (AREA)
  • Computing Systems (AREA)
  • Mathematical Physics (AREA)
  • Pure & Applied Mathematics (AREA)
  • Mathematical Optimization (AREA)
  • Mathematical Analysis (AREA)
  • Computational Mathematics (AREA)
  • Algebra (AREA)
  • Probability & Statistics with Applications (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)
  • Testing And Monitoring For Control Systems (AREA)
  • Debugging And Monitoring (AREA)

Abstract

一种热负荷预测方法及装置，方法包括：设置多种预测状态及对应的误差阈值，形成预测模型（101）；预测模型预测目标蒸汽用户在多个测试时间段内使用锅炉蒸汽时分别对应的第一热负荷量级（102）；根据各个第一热负荷量级确定各个测试时间段分别对应的相对预测误差（103）；根据各相对预测误差形成状态转移概率矩阵，根据状态转移概率矩阵确定各个未来时间段相对于各种预测状态的状态概率（104）；预测模型预测各未来时间段分别对应的第二热负荷量级（105）；针对于每个未来时间段，根据其对应的第二热负荷量级、其相对于各种预测状态的状态概率及各种预测状态分别对应的误差阈值，确定其对应的热负荷区间（106）。该方法可方便后续对锅炉蒸汽进行更为高效的调度。

Description

一种热负荷预测方法、装置、可读介质及电子设备 技术领域
本发明涉及电气工程技术领域，尤其涉及一种热负荷预测方法、装置、可读介质及电子设备。
背景技术
为了实现对蒸汽设备蒸汽进行高效调度，通常需要对蒸汽用户在各个未来时间段内使用蒸汽时所对应的热负荷进行预测。
目前，通常利用自回归积分滑动平均模型(Autoregressive Integrated Moving Average Model,ARIMA模型)对蒸汽用户在各个未来时间段内使用蒸汽设备蒸汽时所分别对应的热负荷量级进行预测。
ARIMA模型仅能对蒸汽用户在各个未来时间段内使用蒸汽时所对应的热负荷量级进行预测时，即预测到一个具体的值，对于在多个连续的历史时间段内无规律使用蒸汽设备蒸汽的蒸汽用户而言，其预测的一个或多个热负荷量级，与该蒸汽用户在相应未来时间段内实际使用蒸汽设备蒸汽时所对应的当前热负荷量级之间可能存在较大的差异，发生单点误差过大现象，导致后续过程中无法对蒸汽设备蒸汽进行高效调度。
发明内容
本发明提供一种热负荷预测方法、装置、可读介质及电子设备，可实现对蒸汽用户在各个未来时间段内使用蒸汽设备蒸汽时所分别对应的热负荷区间进行预测，方便后续过程中对蒸汽设备蒸汽进行更为高效的调度。
第一方面，本发明提供了一种热负荷预测方法，包括：
设置多种预测状态及其分别对应的误差阈值，形成目标蒸汽用户所对应的预测模型；
调用所述预测模型预测所述目标蒸汽用户在多个测试时间段内使用蒸汽设备蒸汽时所分别对应的第一热负荷量级；
根据各个所述测试时间段分别对应的第一热负荷量级确定各个所述测试时间段所分别对应的相对预测误差；
根据各个所述测试时间段所分别对应的相对预测误差及各个所述预测状态所分别对应的误差阈值形成状态转移概率矩阵，并根据所述状态转移概率矩阵确定各个未来时间段相对于各种所述预测状态的状态置信概率；
调用所述预测模型预测所述目标用户在各个所述未来时间段内使用蒸汽设备蒸汽时所分别对应的第二热负荷量级；
针对于每一个所述未来时间段，根据所述未来时间段所对应的第二热负荷量级、所述未来时间段相对于各种所述预测状态的状态置信概率以及各种所述预测状态所分别对应的误差阈值，确定所述目标蒸汽用户在所述未来时间段内使用蒸汽设备蒸汽时所对应的热负荷区间。
优选地，
所述根据各个所述测试时间段分别对应的第一热负荷量级确定各个所述测试时间段所分别对应的相对预测误差，包括：
获取所述目标蒸汽用户在各个所述测试时间段使用蒸汽设备蒸汽时所分别对应的实际热负荷量级；
通过如下公式1计算各个所述测试时间段所分别对应的相对预测误差：
Figure PCTCN2019107928-appb-000001
其中，E i表征第i个测试时间段所对应的相对预测误差、F i表征第i个测试时间段所对应的实际热负荷量级、f i表征第i个测试时间段所对应的第一热负荷量级。
优选地，
所述根据各个所述测试时间段所分别对应的相对预测误差及各个所述预测状态所分别对应的误差阈值形成状态转移概率矩阵，并根据所述状态转移概率矩阵确定各个未来时间段相对于各种所述预测状态的状态置信概率，包括：
根据各个所述测试时间段所分别对应的相对预测误差、各种所述预测状态分别对应对应的误差阈值确定每一个所述测试时间段所分别对应的可信预测状态；
根据每两个相邻的所述测试时间段所分别对应的可信预测状态，确定各种预测状态转移类型所分别对应的状态转移次数；
根据每一种所述状态转移类型分别对应的状态转移次数确定每一种所述状态转移类型所分别对应的状态转移概率；
利用每一种所述状态转移类型分别对应的状态转移概率组成状态转移概率矩阵；
确定当前时间段相对于各种所述预测状态的初始概率矩阵；
根据所述初始概率矩阵及所述状态转移概率矩阵确定各个未来时间段相对于各种所述预测状态的置信概率。
优选地，
所述根据所述初始概率矩阵及所述状态转移概率矩阵确定各个未来时间段相对于各种所述预测状态的置信概率，包括：
通过如下公式2计算各个未来时间段相对于各种所述预测状态的置信概率矩阵：
X(n)＝X(0)*P n     (2)
其中，X(n)表征以当前时间段的截止时间点为起始时间点的第n个未来时间段相对于各种所述预测状态的置信概率矩阵、X(0)表征当前时间段对应的初始概率矩阵、P表征所述状态转移概率矩阵；
针对于每一个所述未来时间段，从所述未来时间段所对应的置信概率矩阵 中提取所述未来时间段相对各种所述预测状态的置信概率。
优选地，
所述根据所述未来时间段所对应的第二热负荷量级、所述未来时间段相对于各种所述预测状态的状态置信概率以及各种所述预测状态所分别对应的误差阈值，确定所述目标蒸汽用户在所述未来时间段内使用蒸汽设备蒸汽时所对应的热负荷区间，包括：
从所述未来时间段相对于各种所述预测状态的置信概率中确定出最大置信概率；
将所述最大置信概率所对应的预测状态确定为目标预测状态；
根据所述目标预测状态所对应的误差阈值以及所述未来时间段所对应的第二热负荷量级确定所述目标蒸汽用户在所述未来时间段内使用蒸汽设备蒸汽时所对应的热负荷区间。
优选地，
所述根据所述目标预测状态所对应的误差阈值以及所述未来时间段所对应的第二热负荷量级确定所述目标蒸汽用户在所述未来时间段内使用蒸汽设备蒸汽时所对应的热负荷区间，包括：
通过如下公式3计算所述目标蒸汽用户在所述未来时间段内使用蒸汽设备蒸汽时所对应的上临界值及临界值：
y＝Y/(1+e)      (3)
其中，y表征所述未来时间段所对应的上临界值或下临界值、Y表所述未来时间段所对应的第二热负荷量级、e表征所述目标预测状态所对应的误差阈值的上极值或下极值；
根据所述上临界值及所述下临界值确定所述目标蒸汽用户在所述未来时间段内使用蒸汽设备蒸汽时所对应的热负荷区间。
优选地，
所述多种预测状态具体包括：极度高估状态、高估状态、正常状态、低估 状态及极度低估状态；其中，
所述极度高估状态所对应的误差阈值具体为大于10％；
所述高估状态所对应的误差阈值具体为大于5％且不大于10％；
所述正常状态所对应的误差阈值具体为不小于-5％且不大于5％；
所述低估状态所对应的误差阈值具体为不小于-10％且小于-5％；
所述极度低估状态所对应的误差阈值具体为小于-10％。
第二方面，本发明提供了一种热负荷预测装置，包括：
预处理模块，用于设置多种预测状态及其分别对应的误差阈值，形成目标蒸汽用户所对应的预测模型；
模型调用模块，用于调用所述预测模型预测所述目标蒸汽用户在多个测试时间段内使用蒸汽设备蒸汽时所分别对应的第一热负荷量级；调用所述预测模型预测所述目标用户在各个所述未来时间段内使用蒸汽设备蒸汽时所分别对应的第二热负荷量级；
误差处理模块，用于根据各个所述测试时间段分别对应的第一热负荷量级确定各个所述测试时间段所分别对应的相对预测误差；
状态概率确定模块，根据各个所述测试时间段所分别对应的相对预测误差及各个所述预测状态所分别对应的误差阈值形成状态转移概率矩阵，并根据所述状态转移概率矩阵确定各个未来时间段相对于各种所述预测状态的状态置信概率；
热负荷区间预测模块，用于针对于每一个所述未来时间段，根据所述未来时间段所对应的第二热负荷量级、所述未来时间段相对于各种所述预测状态的状态置信概率以及各种所述预测状态所分别对应的误差阈值，确定所述目标蒸汽用户在所述未来时间段内使用蒸汽设备蒸汽时所对应的热负荷区间。
第三方面，本发明提供了一种可读介质，包括执行指令，当电子设备的处 理器执行所述执行指令时，所述电子设备执行如第一方面中任一所述的方法。
第四方面，本发明提供了一种电子设备，包括：处理器、存储器和总线；所述存储器用于存储执行指令，所述处理器与所述存储器通过所述总线连接，当所述电子设备运行时，所述处理器执行所述存储器存储的所述执行指令，以使所述处理器执行如第一方面中任一所述的方法。
本发明提供了一种热负荷预测方法、装置、可读介质及电子设备，该方法通过设置多种预测状态及其分别对应的误差阈值，并形成目标蒸汽用户所对应的预测模型，通过调用预测模型预测目标蒸汽用户在多个测试时间段内使用蒸汽设备蒸汽时所分别对应的第一热负荷量级之后，则可根据各个测试时间段分别对应的第一热负荷量级确定各个测试时间段所分别对应的相对预测误差，进而根据各个测试时间段所分别对应的相对预测误差形成状态转移概率矩阵，并根据状态转移概率矩阵确定各个未来时间段相对于各种预测状态的状态概率，再次调用预测模型预测目标用户在各个未来时间段内使用蒸汽设备蒸汽时所分别对应的第二热负荷量级之后，则可针对于每一个未来时间段，根据其对应的第二热负荷量级、其相对于各种预测状态的状态概率以及各种预测状态所分别对应的误差阈值，确定出目标蒸汽用户在该未来时间段内使用蒸汽设备蒸汽时所对应的热负荷区间，实现对预测的第二热负荷量级进行优化而避免发生单点误差过大现象，可方便后续过程中对蒸汽设备蒸汽进行更为高效的调度。
附图说明
为了更清楚地说明本发明实施例或现有的技术方案，下面将对实施例或现有技术描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本发明中记载的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动性的前提下，还可以根据这些附图获得其他的附图。
图1为本发明一实施例提供的一种热负荷预测方法的流程示意图；
图2为本发明一实施例提供的另一种热负荷预测方法的流程示意图；
图3为本发明一实施例提供的一种热负荷预测装置的结构示意图；
图4为本发明一实施例提供的一种电子设备的结构示意图。
具体实施方式
为使本发明的目的、技术方案和优点更加清楚，下面将结合具体实施例及相应的附图对本发明的技术方案进行清楚、完整地描述。显然，所描述的实施例仅是本发明一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。
如图1所示，本发明实施例提供了一种热负荷预测方法，包括：
步骤101，设置多种预测状态及其分别对应的误差阈值，形成目标蒸汽用户所对应的预测模型；
步骤102，调用所述预测模型预测所述目标蒸汽用户在多个测试时间段内使用蒸汽设备蒸汽时所分别对应的第一热负荷量级；
步骤103，根据各个所述测试时间段分别对应的第一热负荷量级确定各个所述测试时间段所分别对应的相对预测误差；
步骤104，根据各个所述测试时间段所分别对应的相对预测误差及各个所述预测状态所分别对应的误差阈值形成状态转移概率矩阵，并根据所述状态转移概率矩阵确定各个未来时间段相对于各种所述预测状态的状态置信概率；
步骤105，调用所述预测模型预测所述目标用户在各个所述未来时间段内使用蒸汽设备蒸汽时所分别对应的第二热负荷量级；
步骤106，针对于每一个所述未来时间段，根据所述未来时间段所对应的第二热负荷量级、所述未来时间段相对于各种所述预测状态的状态置信概率以及各种所述预测状态所分别对应的误差阈值，确定所述目标蒸汽用户在所述未 来时间段内使用蒸汽设备蒸汽时所对应的热负荷区间。
如图1所示的实施例，该方法通过设置多种预测状态及其分别对应的误差阈值，并形成目标蒸汽用户所对应的预测模型，通过调用预测模型预测目标蒸汽用户在多个测试时间段内使用蒸汽设备蒸汽时所分别对应的第一热负荷量级之后，则可根据各个测试时间段分别对应的第一热负荷量级确定各个测试时间段所分别对应的相对预测误差，进而根据各个测试时间段所分别对应的相对预测误差形成状态转移概率矩阵，并根据状态转移概率矩阵确定各个未来时间段相对于各种预测状态的状态概率，再次调用预测模型预测目标用户在各个未来时间段内使用蒸汽设备蒸汽时所分别对应的第二热负荷量级之后，则可针对于每一个未来时间段，根据其对应的第二热负荷量级、其相对于各种预测状态的状态概率以及各种预测状态所分别对应的误差阈值，确定出目标蒸汽用户在该未来时间段内使用蒸汽设备蒸汽时所对应的热负荷区间，实现对预测的第二热负荷量级进行优化而避免发生单点误差过大现象，可方便后续过程中对蒸汽设备蒸汽进行更为高效的调度。
预测模型具体可以是自回归积分滑动平均模型(Auto Regressive Integrated Moving Average Model,ARIMA模型)或GBRT(Gradient Boost Regression Tree，渐进梯度回归树)预测模型，形成ARIMA模型或GBRT预测模型均需要以目标蒸汽用户在多个连续的历史时间段内使用蒸汽设备蒸汽时所分别对应的实际热负荷量级为依据。
需要说明的是，本发明各个实施例中任一处所述的“历史时间段”、“测试时间短”、“未来时间段”的时间长度均相同，其具体可以为1个小时；显而易见的，在一些特殊的业务场景中，时间段的时间长度也可以设置为其它数值。
本发明一个优选实施例中，所述多种预测状态具体包括：极度高估状态、高估状态、正常状态、低估状态及极度低估状态；其中，
所述极度高估状态所对应的误差阈值具体为大于10％；
所述高估状态所对应的误差阈值具体为大于5％且不大于10％；
所述正常状态所对应的误差阈值具体为不小于-5％且不大于5％；
所述低估状态所对应的误差阈值具体为不小于-10％且小于-5％；
所述极度低估状态所对应的误差阈值具体为小于-10％。
该实施例中所设置的多种预测状态，可由用户结合实际业务场景的不同需求增加或减少，各个预测状态分别配置不同的误差阈值，经过了有限次实现进行验证，各个预测状态分别对应的误差阈值为前述实施例中所述的阈值时，可确保目标蒸汽用户在各个未来时间段内使用蒸汽设备蒸汽时所分别对应的实际热负荷量级位于通过本发明实施例提供的技术方案所预测得到的相应热负荷区间内，即确保本发明实施例提供的技术方案能够更为准确的实现对目标蒸汽用户在各个未来时间段内使用蒸汽设备蒸汽时所分别对应的热负荷量级进行预测。
为方便描述，本发明后续各个实施例中仅利用S1表征极度高估状态、利用S2表征高估状态、利用S3表征正常状态、利用S4表征低估状态、利用S5表征极度低估状态。
本发明一个优选实施例中，所述根据各个所述测试时间段分别对应的第一热负荷量级确定各个所述测试时间段所分别对应的相对预测误差，包括：
获取所述目标蒸汽用户在各个所述测试时间段使用蒸汽设备蒸汽时所分别对应的实际热负荷量级；
通过如下公式1计算各个所述测试时间段所分别对应的相对预测误差：
Figure PCTCN2019107928-appb-000002
其中，E i表征第i个测试时间段所对应的相对预测误差、F i表征第i个测 试时间段所对应的实际热负荷量级、f i表征第i个测试时间段所对应的第一热负荷量级。
该实施例中，通过预测模型预测目标蒸汽用户在各个测试时间段内使用蒸汽设备蒸汽时所分别对应的第一热负荷量级，在获取到目标蒸汽用户于各个测试时间段内使用蒸汽设备蒸汽时所分别对应的实际热负荷量级之后，进一步通过公式1对各个测试时间段所分别对应的第一热负荷量级及实际热负荷量级进行处理，得到的每一个相对预测误差则能够用于评价相应测试时间段所对应预测的第一热负荷量级与实际热负荷量级之间的预测状态，并实现对相应测试时间段所对应的第一热负荷量级的误差大小进行度量。
本发明与一个优选实施例中，所述根据各个所述测试时间段所分别对应的相对预测误差及各个所述预测状态所分别对应的误差阈值形成状态转移概率矩阵，并根据所述状态转移概率矩阵确定各个未来时间段相对于各种所述预测状态的状态置信概率，包括：
根据各个所述测试时间段所分别对应的相对预测误差、各种所述预测状态分别对应对应的误差阈值确定每一个所述测试时间段所分别对应的可信预测状态；
根据每两个相邻的所述测试时间段所分别对应的可信预测状态，确定各种预测状态转移类型所分别对应的状态转移次数；
根据每一种所述状态转移类型分别对应的状态转移次数确定每一种所述状态转移类型所分别对应的状态转移概率；
利用每一种所述状态转移类型分别对应的状态转移概率组成状态转移概率矩阵；
确定当前时间段相对于各种所述预测状态的初始概率矩阵；
根据所述初始概率矩阵及所述状态转移概率矩阵确定各个未来时间段相对于各种所述预测状态的置信概率。
举例来说，若通过前述公式1计算出第i个测试时间段所对应相对预测误差E i是6％，根据前述实施例中所述预测状态S1、S2、S3、S4、S5所分别对应的误差阈值，则可确定出6％位于高估状态S2所对应的误差阈值内，那么，则可将高估状态S2确定为第i个测试时间段所对应可信预测状态。
该实施例中，预测状态转移类型具体指的是设置的多种预测状态中，一种预测状态向当前预测状态或其他预测状态发生转移的情况；具体以上述实施例所述的5种预测状态S1、S2、S3、S4、S5为例，针对于S1，其对应有S1向S1转换、S1向S2转换、S1向S转换、S1向S4转换、S4向S5转换共5种转换情况，基于相似的原理，S2、S3、S4、S5也分别对应有5种转换情况，每一种转换情况即为一种预测状态转换类型，那么，这里则可具有25种预测状态转换类型。
该实施例中，以n个连续的测试时间段为例，若第i-1个测试时间段所对应的可信预测状态是S1、第i个测试时间段所对应的可信预测状态是S2，则可将S1向S2转换的预测状态转换类型所对应的状态转移次数加1；针对每两个相邻的测试时间段均进行前述处理之后，则可得到个中预测状态转换类型所分别对应的状态转移次数；进一步的，针对于一种状态转换类型，其对应的状态转移次数与测试时间段总数n的比值即为该中状态转换类型所对应的状态转移概率P。
具体地，各个状态转移类型所分别对应的状态转移概率在形成的状态转移概率矩阵中的位置可参考如下表1：
表1
P S1-S1 P S1-S2 P S1-S3 P S1-S4 P S1-S5
P S2-S1 P S2-S2 P S2-S3 P S2-S4 P S2-S5
P S3-S1 P S3-S2 P S3-S3 P S3-S4 P S3-S5
P S4-S1 P S4-S2 P S4-S3 P S4-S4 P S4-S5
P S5-S1 P S5-S2 P S5-S3 P S5-S4 P S5-S5
需要说明的是，表1中状态转移概率P S1-S2具提表征状态转移类型“S1向S2转换”所对应的状态转移概率，本领域技术人员可依据相似的规则理解表1中记载的各个状态转移概率。
还需要说明的是，每一个状态转移概率在表1中对应的行列位置，即为该状态转移概率作为位于状态转移概率矩阵中的一个元素时，该元素对应在状态转移概率中的行列位置。
该实施例中，确定当前时间段相对于各种预测状态的初始概率矩阵时，可通过相似的方法确定出当前时间段所对应的可信预测状态，初始概率矩阵中具有5个元素，5个元素依次为预测状态S1、S2、S3、S4、S5的初始概率，且仅有当前时间段所对应的可信预测状态的初始概率为1，其他预测状态的初始概率均为0。
该实施例中，根据初始概率矩阵及状态转移概率矩阵确定各个未来时间段相对于各种预测状态的置信概率时，针对于一个未来时间段，该未来时间段相对于一种当前预测状态的置信概率越高，则说明调用形成的预测模型对该目标蒸汽用户在该未来时间段内使用过蒸汽设备蒸汽时所对应的第二热负荷量级进行预测时，预测的第二热负荷量级所对应的预测状态为当前预测状态的可能性越高，从而方便后续过程中根据该未来时间段相对于各种预测状态的置信概率，对通过预测模型预测得到的该未来时间段所对应的第二热负荷量级进行优化，实现更为准确的确定出目标蒸汽用户在该未来时间段内使用蒸汽设备蒸汽时所对应的热负荷区间。
具体地，本发明一个实施例中，所述根据所述初始概率矩阵及所述状态转移概率矩阵确定各个未来时间段相对于各种所述预测状态的置信概率，包括：
通过如下公式2计算各个未来时间段相对于各种所述预测状态的置信概率矩阵：
X(n)＝X(0)*P n    (2)
其中，X(n)表征以当前时间段的截止时间点为起始时间点的第n个未来时间段相对于各种所述预测状态的置信概率矩阵、X(0)表征当前时间段对应的初始概率矩阵、P表征所述状态转移概率矩阵；
针对于每一个所述未来时间段，从所述未来时间段所对应的置信概率矩阵中提取所述未来时间段相对各种所述预测状态的置信概率。
通过前述公式2得到的第n个时间段所对应的置信概率矩阵中，以前述实施例中所述多种预测状态具体包括S1、S2、S3、S4、S5为例，该置信概率矩阵中的则可包括5个元素，5个元素依次为第n个未来时间段相对于预测状态S1、S2、S3、S4、S5的置信概率。
在本发明一个优选实施例中，所述根据所述未来时间段所对应的第二热负荷量级、所述未来时间段相对于各种所述预测状态的状态置信概率以及各种所述预测状态所分别对应的误差阈值，确定所述目标蒸汽用户在所述未来时间段内使用蒸汽设备蒸汽时所对应的热负荷区间，包括：
从所述未来时间段相对于各种所述预测状态的置信概率中确定出最大置信概率；
将所述最大置信概率所对应的预测状态确定为目标预测状态；
根据所述目标预测状态所对应的误差阈值以及所述未来时间段所对应的第二热负荷量级确定所述目标蒸汽用户在所述未来时间段内使用蒸汽设备蒸汽时所对应的热负荷区间。
具体地，本发明一个优选实施例中，所述根据所述目标预测状态所对应的误差阈值以及所述未来时间段所对应的第二热负荷量级确定所述目标蒸汽用户在所述未来时间段内使用蒸汽设备蒸汽时所对应的热负荷区间，包括：
通过如下公式3计算所述目标蒸汽用户在所述未来时间段内使用蒸汽设备蒸汽时所对应的上临界值及临界值：
y＝Y/(1+e)    (3)
其中，y表征所述未来时间段所对应的上临界值或下临界值、Y表所述未来时间段所对应的第二热负荷量级、e表征所述目标预测状态所对应的误差阈值的上极值或下极值；
根据所述上临界值及所述下临界值确定所述目标蒸汽用户在所述未来时间段内使用蒸汽设备蒸汽时所对应的热负荷区间。
该实施例中，若一个未来时间段相对于某一种当前预测状态的置信概率越高，则说明调用预测模型对该目标蒸汽用户在该未来时间段内使用蒸汽设备蒸汽时所对应的第二热负荷量级进行预测时，预测的第二热负荷量级所对应的预测状态为当前预测状态的可能性越高；因此，将最大置信概率所对应的预测状态确定为目标状态之后，根据该目标预测状态所对应的误差阈值，通过前述公式3对预测的第二热负荷量级进行优化，则可实现更为准确的确定出目标蒸汽用户在该未来时间段内使用蒸汽设备蒸汽时所对应的热负荷区间。
举例来说，若确定出第j个未来时间段相对于前述预测状态S1、S2、S3、S4、S5的各个状态置信概率中，对应于预测状态S2的置信概率最高，则说明通过预测模型对第j个未来时间段内用户使用蒸汽设备蒸汽时所对应的第二热负荷量级进行预测时，预测的第二热负荷量级对应为高估状态的可能性越高；具体地，参见前述实施例中预测状态S2所对应的误差阈值时大于5％且不大于10％，即高估状态S2所对应的上极值(最大值或最大值所趋近的极限值)为10％，下极限值(最小值或最小值所趋近的极限值)为5％，通过预测模型预测的第j个未来时间段所对应的第二热负荷量级是Y(预测的Y值发生高估的概率极高)，那么，通过上述公式3则可计算出第i个未来时间段所对应的上临界值为0.95Y，且下临界值为0.91Y，进而确定出第i个未来时间段内用户使用蒸汽设备蒸汽时所对应的热负荷区间为0.91Y～0.95Y。
基于前述举例，后续过程中针对蒸汽进行调度时，若根据预测模型预测的的第二热负荷量级Y在第i个未来时间段内对目标蒸汽用户进行蒸汽调度，则可能导致在目标蒸汽用户出发生供需差异过大，反之，根据热负荷区间 0.91Y～0.95Y在第i个未来时间段内对目标蒸汽用户进行蒸汽调度时，表征目标蒸汽用户的最低热负荷需求可能为0.91Y、最高热负荷需求可能为0.95Y、热负荷需求中间值为0.9Y，此时则可结合实际业务场景以目标蒸汽用户的最低热负荷需求、最高热负荷需求或热负荷需求中间值对目标蒸汽用户进行蒸汽调度；不难看出，通过本发明实施例提供的技术方案，可克服通过预测模型直接预测未来时间段内用户使用蒸汽设备蒸汽时所对应的热负荷量级而造成的单点误差过大现象，可方便后续过程中对蒸汽设备蒸汽进行更为高效的调度。
请参考图2，本发明实施例提供了另一种热负荷预测方法，具体包括如下各个步骤。
步骤201，设置多种预测状态及其分别对应的误差阈值，以及根据目标用户在多个连续的历史时间段内使用蒸汽设备蒸汽时所分别对应的历史热负荷量级形成预测模型。
步骤202，调用预测模型预测目标蒸汽用户在多个连续的测试时间段内使用蒸汽设备蒸汽时所分别对应的第一热负荷量级。
步骤203，获取目标蒸汽用户在各个测试时间段使用蒸汽设备蒸汽时所分别对应的实际热负荷量级。
步骤204，计算各个测试时间段所分别对应的相对预测误差。
步骤204可具体结合本发明任意一个实施例中提供的公式1实现。
步骤205，根据各个测试时间段所分别对应的相对预测误差、各种预测状态分别对应对应的误差阈值确定每一个测试时间段所分别对应的可信预测状态。
步骤206，根据每两个相邻的测试时间段所分别对应的可信预测状态，确定各种预测状态转移类型所分别对应的状态转移次数。
步骤207，根据每一种状态转移类型分别对应的状态转移次数确定每一种状态转移类型所分别对应的状态转移概率。
步骤208，利用每一种状态转移类型所分别对应的状态转移概率组成状态转移概率矩阵。
步骤209，确定当前时间段相对于各种预测状态的初始概率矩阵。
步骤210，计算各个未来时间段相对于各种预测状态的置信概率矩阵。
步骤210可具体结合本发明任意一个实施例中提供的公式2实现。
步骤211，针对于每一个未来时间段，从未来时间段所对应的置信概率矩阵中提取未来时间段相对各种预测状态的置信概率。
为方便描述，后续各步骤仅针对一个未来时间段进行相应的处理以实现确定目标蒸汽用户在该未来时间段内使用蒸汽设备蒸汽时所对应的热负荷区间。
步骤212，从未来时间段相对于各种预测状态的置信概率中确定出最大置信概率，并将最大置信概率所对应的预测状态确定为目标预测状态。
步骤213，根据目标预测状态所对应的误差阈值以及未来时间段所对应的第二热负荷量级确定目标蒸汽用户在未来时间段内使用蒸汽设备蒸汽时所对应的热负荷区间。
这里，具体可结合上述前述任一实施例中提供的公式3实现确定目标蒸汽用户在未来时间段内使用蒸汽设备蒸汽时所对应的热负荷区间。
基于与本发明方法实施例相同的构思，如图3所示，本发明实施例还提供了一种热负荷预测装置，包括：
预处理模块301，用于设置多种预测状态及其分别对应的误差阈值，形成目标蒸汽用户所对应的预测模型；
模型调用模块302，用于调用所述预测模型预测所述目标蒸汽用户在多个测试时间段内使用蒸汽设备蒸汽时所分别对应的第一热负荷量级；调用所述预测模型预测所述目标用户在各个所述未来时间段内使用蒸汽设备蒸汽时所分别对应的第二热负荷量级；
误差处理模块303，用于根据各个所述测试时间段分别对应的第一热负荷 量级确定各个所述测试时间段所分别对应的相对预测误差；
状态概率确定模块304，根据各个所述测试时间段所分别对应的相对预测误差及各个所述预测状态所分别对应的误差阈值形成状态转移概率矩阵，并根据所述状态转移概率矩阵确定各个未来时间段相对于各种所述预测状态的状态置信概率；
热负荷区间预测模块305，用于针对于每一个所述未来时间段，根据所述未来时间段所对应的第二热负荷量级、所述未来时间段相对于各种所述预测状态的状态置信概率以及各种所述预测状态所分别对应的误差阈值，确定所述目标蒸汽用户在所述未来时间段内使用蒸汽设备蒸汽时所对应的热负荷区间。
图4是本发明的一个实施例电子设备的结构示意图。在硬件层面，该电子设备包括处理器，可选地还包括内部总线、网络接口、存储器。其中，存储器可能包含内存，例如高速随机存取存储器(Random-Access Memory，RAM)，也可能还包括非易失性存储器(non-volatile memory)，例如至少1个磁盘存储器等。当然，该电子设备还可能包括其他业务所需要的硬件。
处理器、网络接口和存储器可以通过内部总线相互连接，该内部总线可以是ISA(Industry Standard Architecture，工业标准体系结构)总线、PCI(Peripheral Component Interconnect，外设部件互连标准)总线或EISA(Extended Industry Standard Architecture，扩展工业标准结构)总线等。所述总线可以分为地址总线、数据总线、控制总线等。为便于表示，图4中仅用一个双向箭头表示，但并不表示仅有一根总线或一种类型的总线。
存储器，用于存放程序/执行指令。具体地，程序可以包括程序代码，所述程序代码包括计算机操作指令。存储器可以包括内存和非易失性存储器，并向处理器提供指令和数据。
在一种可能实现的方式中，处理器从非易失性存储器中读取对应的计算机程序到内存中然后运行，也可从其它设备上获取相应的计算机程序，以在逻辑 层面上形成热负荷预测装置。处理器，执行存储器所存放的程序，以通过执行的程序实现本发明任一实施例中提供的热负荷预测方法。
上述实施例提供的热负荷预测装置执行的方法可以应用于处理器中，或者由处理器实现。处理器可能是一种集成电路芯片，具有信号的处理能力。在实现过程中，上述方法的各步骤可以通过处理器中的硬件的集成逻辑电路或者软件形式的指令完成。上述的处理器可以是通用处理器，包括中央处理器(Central Processing Unit，CPU)、网络处理器(Network Processor，NP)等；还可以是数字信号处理器(Digital Signal Processor，DSP)、专用集成电路(Application Specific Integrated Circuit，ASIC)、现场可编程门阵列(Field－Programmable Gate Array，FPGA)或者其他可编程逻辑器件、分立门或者晶体管逻辑器件、分立硬件组件。可以实现或者执行本发明实施例中的公开的各方法、步骤及逻辑框图。通用处理器可以是微处理器或者该处理器也可以是任何常规的处理器等。
结合本发明实施例所公开的方法的步骤可以直接体现为硬件译码处理器执行完成，或者用译码处理器中的硬件及软件模块组合执行完成。软件模块可以位于随机存储器，闪存、只读存储器，可编程只读存储器或者电可擦写可编程存储器、寄存器等本领域成熟的存储介质中。该存储介质位于存储器，处理器读取存储器中的信息，结合其硬件完成上述方法的步骤。
本发明实施例还提出了一种计算机可读存储介质，该计算机可读存储介质存储一个或多个程序，该一个或多个程序包括指令，该指令当被包括多个应用程序的电子设备执行时，能够使该电子设备执行本发明任一实施例中提供的热负荷预测方法，并具体用于执行如图1和/或图2所示的方法。
上述实施例阐明的系统、装置、模块或单元，具体可以由计算机芯片或实体实现，或者由具有某种功能的产品来实现。一种典型的实现设备为计算机。具体的，计算机例如可以为个人计算机、膝上型计算机、蜂窝电话、相机电话、智能电话、个人数字助理、媒体播放器、导航设备、电子邮件设备、游戏控制 台、平板计算机、可穿戴设备或者这些设备中的任何设备的组合。
为了描述的方便，描述以上装置时以功能分为各种单元或模块分别描述。当然，在实施本发明时可以把各单元或模块的功能在同一个或多个软件和/或硬件中实现。
本领域内的技术人员应明白，本发明的实施例可提供为方法、系统、或计算机程序产品。因此，本发明可采用完全硬件实施例、完全软件实施例、或结合软件和硬件方面的实施例的形式。而且，本发明可采用在一个或多个其中包含有计算机可用程序代码的计算机可用存储介质(包括但不限于磁盘存储器、CD-ROM、光学存储器等)上实施的计算机程序产品的形式。
本发明是参照根据本发明实施例的方法、设备(系统)、和计算机程序产品的流程图和/或方框图来描述的。应理解可由计算机程序指令实现流程图和/或方框图中的每一流程和/或方框、以及流程图和/或方框图中的流程和/或方框的结合。可提供这些计算机程序指令到通用计算机、专用计算机、嵌入式处理机或其他可编程数据处理设备的处理器以产生一个机器，使得通过计算机或其他可编程数据处理设备的处理器执行的指令产生用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的装置。
这些计算机程序指令也可存储在能引导计算机或其他可编程数据处理设备以特定方式工作的计算机可读存储器中，使得存储在该计算机可读存储器中的指令产生包括指令装置的制造品，该指令装置实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能。
这些计算机程序指令也可装载到计算机或其他可编程数据处理设备上，使得在计算机或其他可编程设备上执行一系列操作步骤以产生计算机实现的处理，从而在计算机或其他可编程设备上执行的指令提供用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的步骤。
在一个典型的配置中，计算设备包括一个或多个处理器(CPU)、输入/输出接口、网络接口和内存。
内存可能包括计算机可读介质中的非永久性存储器，随机存取存储器(RAM)和/或非易失性内存等形式，如只读存储器(ROM)或闪存(flash RAM)。内存是计算机可读介质的示例。
计算机可读介质包括永久性和非永久性、可移动和非可移动媒体可以由任何方法或技术来实现信息存储。信息可以是计算机可读指令、数据结构、程序的模块或其他数据。计算机的存储介质的例子包括，但不限于相变内存(PRAM)、静态随机存取存储器(SRAM)、动态随机存取存储器(DRAM)、其他类型的随机存取存储器(RAM)、只读存储器(ROM)、电可擦除可编程只读存储器(EEPROM)、快闪记忆体或其他内存技术、只读光盘只读存储器(CD-ROM)、数字多功能光盘(DVD)或其他光学存储、磁盒式磁带，磁带磁磁盘存储或其他磁性存储设备或任何其他非传输介质，可用于存储可以被计算设备访问的信息。按照本文中的界定，计算机可读介质不包括暂存电脑可读媒体(transitory media)，如调制的数据信号和载波。
还需要说明的是，术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含，从而使得包括一系列要素的过程、方法、商品或者设备不仅包括那些要素，而且还包括没有明确列出的其他要素，或者是还包括为这种过程、方法、商品或者设备所固有的要素。在没有更多限制的情况下，由语句“包括一个……”限定的要素，并不排除在包括所述要素的过程、方法、商品或者设备中还存在另外的相同要素。
本领域技术人员应明白，本发明的实施例可提供为方法、系统或计算机程序产品。因此，本发明可采用完全硬件实施例、完全软件实施例或结合软件和硬件方面的实施例的形式。而且，本发明可采用在一个或多个其中包含有计算机可用程序代码的计算机可用存储介质(包括但不限于磁盘存储器、CD-ROM、光学存储器等)上实施的计算机程序产品的形式。
本发明可以在由计算机执行的计算机可执行指令的一般上下文中描述，例如程序模块。一般地，程序模块包括执行特定任务或实现特定抽象数据类型的 例程、程序、对象、组件、数据结构等等。也可以在分布式计算环境中实践本发明，在这些分布式计算环境中，由通过通信网络而被连接的远程处理设备来执行任务。在分布式计算环境中，程序模块可以位于包括存储设备在内的本地和远程计算机存储介质中。
本发明中的各个实施例均采用递进的方式描述，各个实施例之间相同相似的部分互相参见即可，每个实施例重点说明的都是与其他实施例的不同之处。尤其，对于系统实施例而言，由于其基本相似于方法实施例，所以描述的比较简单，相关之处参见方法实施例的部分说明即可。
以上所述仅为本发明的实施例而已，并不用于限制本发明。对于本领域技术人员来说，本发明可以有各种更改和变化。凡在本发明的精神和原理之内所作的任何修改、等同替换、改进等，均应包含在本发明的权利要求范围之内。

Claims (10)

  1. 一种热负荷预测方法，其特征在于，包括：
    设置多种预测状态及其分别对应的误差阈值，形成目标蒸汽用户所对应的预测模型；
    调用所述预测模型预测所述目标蒸汽用户在多个测试时间段内使用蒸汽设备蒸汽时所分别对应的第一热负荷量级；
    根据各个所述测试时间段分别对应的第一热负荷量级确定各个所述测试时间段所分别对应的相对预测误差；
    根据各个所述测试时间段所分别对应的相对预测误差及各个所述预测状态所分别对应的误差阈值形成状态转移概率矩阵，并根据所述状态转移概率矩阵确定各个未来时间段相对于各种所述预测状态的状态置信概率；
    调用所述预测模型预测所述目标用户在各个所述未来时间段内使用蒸汽设备蒸汽时所分别对应的第二热负荷量级；
    针对于每一个所述未来时间段，根据所述未来时间段所对应的第二热负荷量级、所述未来时间段相对于各种所述预测状态的状态置信概率以及各种所述预测状态所分别对应的误差阈值，确定所述目标蒸汽用户在所述未来时间段内使用蒸汽设备蒸汽时所对应的热负荷区间。
  2. 根据权利要求1所述的方法，其特征在于，
    所述根据各个所述测试时间段分别对应的第一热负荷量级确定各个所述测试时间段所分别对应的相对预测误差，包括：
    获取所述目标蒸汽用户在各个所述测试时间段使用蒸汽设备蒸汽时所分别对应的实际热负荷量级；
    通过如下公式1计算各个所述测试时间段所分别对应的相对预测误差：
    Figure PCTCN2019107928-appb-100001
    其中，E i表征第i个测试时间段所对应的相对预测误差、F i表征第i个测 试时间段所对应的实际热负荷量级、f i表征第i个测试时间段所对应的第一热负荷量级。
  3. 根据权利要求2所述的方法，其特征在于，
    所述根据各个所述测试时间段所分别对应的相对预测误差及各个所述预测状态所分别对应的误差阈值形成状态转移概率矩阵，并根据所述状态转移概率矩阵确定各个未来时间段相对于各种所述预测状态的状态置信概率，包括：
    根据各个所述测试时间段所分别对应的相对预测误差、各种所述预测状态分别对应的误差阈值确定每一个所述测试时间段所分别对应的可信预测状态；
    根据每两个相邻的所述测试时间段所分别对应的可信预测状态，确定各种预测状态转移类型所分别对应的状态转移次数；
    根据每一种所述状态转移类型分别对应的状态转移次数确定每一种所述状态转移类型所分别对应的状态转移概率；
    利用每一种所述状态转移类型分别对应的状态转移概率组成状态转移概率矩阵；
    确定当前时间段相对于各种所述预测状态的初始概率矩阵；
    根据所述初始概率矩阵及所述状态转移概率矩阵确定各个未来时间段相对于各种所述预测状态的置信概率。
  4. 根据权利要求3所述的方法，其特征在于，
    所述根据所述初始概率矩阵及所述状态转移概率矩阵确定各个未来时间段相对于各种所述预测状态的置信概率，包括：
    通过如下公式2计算各个未来时间段相对于各种所述预测状态的置信概率矩阵：
    X(n)＝X(0)*P n         (2)
    其中，X(n)表征以当前时间段的截止时间点为起始时间点的第n个未来时间段相对于各种所述预测状态的置信概率矩阵、X(0)表征当前时间段对应的初 始概率矩阵、P表征所述状态转移概率矩阵；
    针对于每一个所述未来时间段，从所述未来时间段所对应的置信概率矩阵中提取所述未来时间段相对各种所述预测状态的置信概率。
  5. 根据权利要求1所述的方法，其特征在于，
    所述根据所述未来时间段所对应的第二热负荷量级、所述未来时间段相对于各种所述预测状态的状态置信概率以及各种所述预测状态所分别对应的误差阈值，确定所述目标蒸汽用户在所述未来时间段内使用蒸汽设备蒸汽时所对应的热负荷区间，包括：
    从所述未来时间段相对于各种所述预测状态的置信概率中确定出最大置信概率；
    将所述最大置信概率所对应的预测状态确定为目标预测状态；
    根据所述目标预测状态所对应的误差阈值以及所述未来时间段所对应的第二热负荷量级确定所述目标蒸汽用户在所述未来时间段内使用蒸汽设备蒸汽时所对应的热负荷区间。
  6. 根据权利要求5所述的方法，其特征在于，
    所述根据所述目标预测状态所对应的误差阈值以及所述未来时间段所对应的第二热负荷量级确定所述目标蒸汽用户在所述未来时间段内使用蒸汽设备蒸汽时所对应的热负荷区间，包括：
    通过如下公式3计算所述目标蒸汽用户在所述未来时间段内使用蒸汽设备蒸汽时所对应的上临界值及临界值：
    y＝Y/(1+e)        (3)
    其中，y表征所述未来时间段所对应的上临界值或下临界值、Y表所述未来时间段所对应的第二热负荷量级、e表征所述目标预测状态所对应的误差阈值的上极值或下极值；
    根据所述上临界值及所述下临界值确定所述目标蒸汽用户在所述未来时间段内使用蒸汽设备蒸汽时所对应的热负荷区间。
  7. 根据权利要求1至6中任一所述的方法，其特征在于，
    所述多种预测状态具体包括：极度高估状态、高估状态、正常状态、低估状态及极度低估状态；其中，
    所述极度高估状态所对应的误差阈值具体为大于10％；
    所述高估状态所对应的误差阈值具体为大于5％且不大于10％；
    所述正常状态所对应的误差阈值具体为不小于-5％且不大于5％；
    所述低估状态所对应的误差阈值具体为不小于-10％且小于-5％；
    所述极度低估状态所对应的误差阈值具体为小于-10％。
  8. 一种热负荷预测装置，其特征在于，包括：
    预处理模块，用于设置多种预测状态及其分别对应的误差阈值，形成目标蒸汽用户所对应的预测模型；
    模型调用模块，用于调用所述预测模型预测所述目标蒸汽用户在多个测试时间段内使用蒸汽设备蒸汽时所分别对应的第一热负荷量级；调用所述预测模型预测所述目标用户在各个所述未来时间段内使用蒸汽设备蒸汽时所分别对应的第二热负荷量级；
    误差处理模块，用于根据各个所述测试时间段分别对应的第一热负荷量级确定各个所述测试时间段所分别对应的相对预测误差；
    状态概率确定模块，根据各个所述测试时间段所分别对应的相对预测误差及各个所述预测状态所分别对应的误差阈值形成状态转移概率矩阵，并根据所述状态转移概率矩阵确定各个未来时间段相对于各种所述预测状态的状态置信概率；
    热负荷区间预测模块，用于针对于每一个所述未来时间段，根据所述未来时间段所对应的第二热负荷量级、所述未来时间段相对于各种所述预测状态的状态置信概率以及各种所述预测状态所分别对应的误差阈值，确定所述目标蒸汽用户在所述未来时间段内使用蒸汽设备蒸汽时所对应的热负荷区间。
  9. 一种可读介质，包括执行指令，当电子设备的处理器执行所述执行指 令时，所述电子设备执行如权利要求1至8中任一所述的方法。
  10. 一种电子设备，包括：处理器、存储器和总线；所述存储器用于存储执行指令，所述处理器与所述存储器通过所述总线连接，当所述电子设备运行时，所述处理器执行所述存储器存储的所述执行指令，以使所述处理器执行如权利要求1至8中任一所述的方法。
PCT/CN2019/107928 2018-09-25 2019-09-25 一种热负荷预测方法、装置、可读介质及电子设备 WO2020063686A1 (zh)

Priority Applications (4)

Application Number Priority Date Filing Date Title
EP19865708.2A EP3822839A4 (en) 2018-09-25 2019-09-25 THERMAL LOAD PREDICTING PROCESS AND APPARATUS, READABLE SUPPORT, AND ELECTRONIC DEVICE
US17/058,137 US11887020B2 (en) 2018-09-25 2019-09-25 Thermal load prediction method and apparatus, readable medium, and electronic device
SG11202102668YA SG11202102668YA (en) 2018-09-25 2019-09-25 Thermal load prediction method and apparatus, readable medium, and electronic device
JP2021514404A JP7426992B2 (ja) 2018-09-25 2019-09-25 熱負荷予測方法、装置、読み取り可能な媒体及び電子機器

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201811113972.6A CN109117595B (zh) 2018-09-25 2018-09-25 一种热负荷预测方法、装置、可读介质及电子设备
CN201811113972.6 2018-09-25

Publications (1)

Publication Number Publication Date
WO2020063686A1 true WO2020063686A1 (zh) 2020-04-02

Family

ID=64856659

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/107928 WO2020063686A1 (zh) 2018-09-25 2019-09-25 一种热负荷预测方法、装置、可读介质及电子设备

Country Status (6)

Country Link
US (1) US11887020B2 (zh)
EP (1) EP3822839A4 (zh)
JP (1) JP7426992B2 (zh)
CN (1) CN109117595B (zh)
SG (1) SG11202102668YA (zh)
WO (1) WO2020063686A1 (zh)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113591380A (zh) * 2021-07-28 2021-11-02 浙江大学 基于图高斯过程的交通流预测方法、介质及设备
CN114518911A (zh) * 2022-02-21 2022-05-20 中国农业银行股份有限公司 一种插件加载时长预测方法、装置、设备和存储介质
CN116432554A (zh) * 2023-04-18 2023-07-14 南方电网电力科技股份有限公司 一种锅炉水动力计算方法、系统、设备和介质

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109117595B (zh) * 2018-09-25 2021-06-25 新智数字科技有限公司 一种热负荷预测方法、装置、可读介质及电子设备
CN110008623B (zh) * 2019-04-15 2023-06-27 廊坊新奥泛能网络科技服务有限公司 蒸汽项目的供能设备选取方法、装置及电子设备
CN113486583B (zh) * 2021-07-06 2024-01-12 新奥新智科技有限公司 设备的健康评估方法、装置、计算机设备及计算机可读存储介质
CN114516403B (zh) * 2022-03-07 2023-05-05 中国空气动力研究与发展中心低速空气动力研究所 一种基于光纤结冰探测的电热防除冰系统功率控制方法
CN115983114A (zh) * 2022-12-19 2023-04-18 广东邦普循环科技有限公司 一种窑炉温度预测方法、系统、设备及介质

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103400204A (zh) * 2013-07-26 2013-11-20 华南理工大学 基于svm-马尔科夫组合法的太阳能光伏发电量预测方法
CN104850916A (zh) * 2015-05-31 2015-08-19 上海电机学院 一种改进型灰色马尔可夫模型的电力设备故障预测方法
CN106485348A (zh) * 2016-09-22 2017-03-08 中国银联股份有限公司 一种交易数据的预测方法及装置
US20170278116A1 (en) * 2016-03-28 2017-09-28 International Business Machines Corporation Forecast model calibration
CN109117595A (zh) * 2018-09-25 2019-01-01 新智数字科技有限公司 一种热负荷预测方法、装置、可读介质及电子设备

Family Cites Families (18)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH09204218A (ja) * 1996-01-24 1997-08-05 Hitachi Ltd プラント監視装置
US8180642B2 (en) * 2007-06-01 2012-05-15 Xerox Corporation Factorial hidden Markov model with discrete observations
KR101644673B1 (ko) 2009-12-15 2016-08-01 램 리써치 코포레이션 Ｃｄ 균일성을 향상시키기 위한 기판 온도의 조절
CN103530527A (zh) * 2013-10-30 2014-01-22 国家电网公司 基于数值天气预报集合预报结果的风电功率概率预测方法
JP6467953B2 (ja) * 2015-01-30 2019-02-13 中国電力株式会社 気温予測システム、気温予測方法およびプログラム
WO2016178316A1 (ja) * 2015-05-07 2016-11-10 日本電気株式会社 計算機調達予測装置、計算機調達予測方法、及び、プログラム
CN105337760A (zh) * 2015-09-21 2016-02-17 浪潮集团有限公司 一种资源使用情况的预测方法及装置
CN105512745A (zh) * 2015-11-05 2016-04-20 天津大学 基于粒子群—bp神经网络的风电功率区间预测方法
CN106845663A (zh) * 2015-12-07 2017-06-13 中国电力科学研究院 一种基于数值天气预报进行热负荷预测的方法
CN105825040A (zh) * 2015-12-29 2016-08-03 海南电力技术研究院 短期电力负荷预测方法
CN105956682A (zh) * 2016-04-19 2016-09-21 上海电力学院 基于bp神经网络与马尔可夫链的短期电价预测方法
CA2965923A1 (en) * 2016-05-05 2017-11-05 Minds Inc. Vehicle load temperature monitoring system
JP6655182B2 (ja) * 2016-06-13 2020-02-26 株式会社日立製作所 空調機管理装置、および、空調機管理方法
US10310852B2 (en) * 2016-09-29 2019-06-04 Entit Software Llc Timing estimations for application lifecycle management work items determined through machine learning
CN106777487B (zh) 2016-11-18 2019-11-05 清华大学 一种含有储能系统的光伏电站可信容量计算方法及系统
CN108461150A (zh) * 2017-02-20 2018-08-28 天津工业大学 一种职业卫生预测研究方法
CN107358060B (zh) * 2017-09-06 2019-10-11 大连理工大学 一种对风电功率预测误差区间进行估计的方法
CN108052387B (zh) 2017-11-15 2021-07-27 青海大学 一种移动云计算中的资源分配预测方法及系统

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103400204A (zh) * 2013-07-26 2013-11-20 华南理工大学 基于svm-马尔科夫组合法的太阳能光伏发电量预测方法
CN104850916A (zh) * 2015-05-31 2015-08-19 上海电机学院 一种改进型灰色马尔可夫模型的电力设备故障预测方法
US20170278116A1 (en) * 2016-03-28 2017-09-28 International Business Machines Corporation Forecast model calibration
CN106485348A (zh) * 2016-09-22 2017-03-08 中国银联股份有限公司 一种交易数据的预测方法及装置
CN109117595A (zh) * 2018-09-25 2019-01-01 新智数字科技有限公司 一种热负荷预测方法、装置、可读介质及电子设备

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP3822839A4 *

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113591380A (zh) * 2021-07-28 2021-11-02 浙江大学 基于图高斯过程的交通流预测方法、介质及设备
CN113591380B (zh) * 2021-07-28 2022-03-22 浙江大学 基于图高斯过程的交通流预测方法、介质及设备
CN114518911A (zh) * 2022-02-21 2022-05-20 中国农业银行股份有限公司 一种插件加载时长预测方法、装置、设备和存储介质
CN114518911B (zh) * 2022-02-21 2024-03-12 中国农业银行股份有限公司 一种插件加载时长预测方法、装置、设备和存储介质
CN116432554A (zh) * 2023-04-18 2023-07-14 南方电网电力科技股份有限公司 一种锅炉水动力计算方法、系统、设备和介质

Also Published As

Publication number Publication date
JP2022500768A (ja) 2022-01-04
CN109117595B (zh) 2021-06-25
SG11202102668YA (en) 2021-04-29
JP7426992B2 (ja) 2024-02-02
EP3822839A1 (en) 2021-05-19
CN109117595A (zh) 2019-01-01
US11887020B2 (en) 2024-01-30
EP3822839A4 (en) 2021-09-01
US20210304045A1 (en) 2021-09-30

Similar Documents

Publication Publication Date Title
WO2020063686A1 (zh) 一种热负荷预测方法、装置、可读介质及电子设备
EP3576036A1 (en) Service execution method and device
TWI709931B (zh) 一種指標異常檢測方法、裝置以及電子設備
WO2020093637A1 (zh) 设备状态预测方法、系统、计算机装置及存储介质
US8578023B2 (en) Computer resource utilization modeling for multiple workloads
CN106293914B (zh) 一种任务调度的方法及终端
CN109857984B (zh) 一种锅炉负荷率-效能曲线的回归方法和装置
CN108073446A (zh) 超时预判方法及装置
Li et al. Siphon extraction for deadlock control in flexible manufacturing systems by using Petri nets
CN108804719B (zh) 数字电路老化仿真方法及装置
US20220138557A1 (en) Deep Hybrid Graph-Based Forecasting Systems
CN110993014B (zh) Ssd处于空闲状态下的行为测试方法、装置、计算机设备及存储介质
CN108470242B (zh) 风险管控方法、装置、及服务器
WO2019192262A1 (zh) 一种评价商户的经营状况的方法、装置及设备
CN112685390B (zh) 数据库实例管理方法及装置、计算设备
KR101674324B1 (ko) 실시간 제어 응용에 적용되는 태스크 스케쥴링 장치 및 방법
WO2020164275A1 (zh) 基于预测模型的处理结果预测方法、装置及服务器
CN109615465B (zh) 业务订单的处理方法、装置和电子设备
CN114186123A (zh) 用于热点事件的处理方法、装置、设备及存储介质
WO2019227415A1 (zh) 一种评分卡模型的调整方法、装置、服务器及存储介质
WO2020103027A1 (zh) 网络功耗调整方法、网络功耗调整装置及移动终端
CN109753405B (zh) 应用资源消耗的检测方法、装置、存储介质和电子设备
TW202032484A (zh) 模型合併方法和裝置
CN110209690B (zh) 数据获取方法、装置及电子设备
CN114610575B (zh) 应用于计算分支的更新峰值的方法、装置、设备和介质

Legal Events

Date Code Title Description
ENP Entry into the national phase

Ref document number: 2021514404

Country of ref document: JP

Kind code of ref document: A

NENP Non-entry into the national phase

Ref country code: DE