Multi-machine linkage device integration energy-saving management and intelligent decision method and system

Abstract

The present application relates to the technical field of equipment energy-saving management, and discloses a multi-machine linkage equipment integrated energy-saving management and intelligent decision method and system, comprising: step 1, obtaining equipment operation parameters, calculating unit energy consumption, determining a runnable load interval, and splitting a task into standard task units; step 2, determining group-level remaining available energy consumption and establishing output and energy consumption constraints; step 3, according to unit energy consumption and in combination with equipment energy consumption coefficients, iteratively obtaining final throughput targets of each device; step 4, calculating marginal energy consumption cost and allocating standard task units to obtain final task allocation; step 5, generating a strategy parameter set based on the final task allocation and the throughput target and monitoring an out-of-limit state, updating the strategy and issuing an alarm when the out-of-limit state occurs; and step 6, generating energy consumption settlement and management output based on actual energy consumption and energy consumption budget. The present application realizes intelligent management of controllable energy consumption budget, accountable resource allocation and evaluatable energy-saving performance in a multi-machine collaborative scenario.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of energy management and resource allocation, and particularly relates to a multi-machine linkage device integrated energy-saving management and intelligent decision method and system. BACKGROUND

[0002] With the evolution of data centers, intelligent manufacturing lines and park-level energy systems towards scale and platform, enterprises usually carry multiple business and multi-shift production and computing tasks in the form of device groups or machine groups. In such scenarios, energy consumption has expanded from single-machine operation cost to an integrated management object that needs budgeting, quota allocation, execution accounting and performance evaluation. In the prior art, one type of solution focuses on energy-saving control or single-point optimization at the device operation level, often limiting consumption with fixed thresholds, empirical coefficients or static quotas, lacking group-level budget account and dynamic constraint modeling for time slots, and being difficult to maintain the balance between budget compliance and output achievement when task demand fluctuates. Another type of solution focuses on energy consumption statistics and report display, usually only summarizing historical energy consumption, lacking association with task plan volume, throughput target and device operable load range, resulting in difficulty in timely locating budget deviation, non-uniform cost allocation, and insufficient fairness across devices. Especially in multi-device parallel collaboration scenarios, if the resource allocation basis is not established based on unit energy consumption, marginal energy consumption cost and budget utilization rate, local device over-consumption may occur while the overall budget still has a surplus, or the overall budget may be exceeded while the responsibility boundary is unclear, making it difficult for energy accounting, budget execution auditing and energy-saving performance evaluation. SUMMARY

[0003] The application provides a multi-machine linkage device integrated energy-saving management and intelligent decision method and system, which solves the technical problems in the related art that energy consumption budget and task allocation lack unified management caliber in a multi-device scenario, budget execution is difficult to quantify and trace, responsibility boundaries are not clear due to unclear energy consumption cost allocation rules, and energy-saving performance evaluation and alarm mechanisms are difficult to be linked with time slot level strategy parameters.

[0004] The application provides a multi-machine linkage device integrated energy-saving management and intelligent decision method, which comprises the following steps:

[0005] Step 1, obtaining the operating parameters of each device, calculating the current unit energy consumption, determining the operable load range of each device, and splitting the to-be-processed task into standard task units;

[0006] Step 2, obtaining the group-level remaining available energy consumption in the current time slot according to the group-level remaining available energy consumption in the previous time slot, the actual energy consumption of each device in the previous time slot and the group-level newly available energy consumption in the current time slot, and establishing group-level output constraints and group-level energy consumption constraints;

[0007] Step 3, initial throughput is generated according to current unit energy consumption, and iterative adjustment is carried out in combination with group-level output constraint, group-level energy consumption constraint and set device energy consumption coefficient, to obtain final throughput targets of each device;

[0008] Step 4, marginal energy consumption cost is calculated based on final throughput targets of each device and device energy consumption coefficient, devices are sorted and standard task units are allocated, to obtain final task allocation of each device;

[0009] Step 5, strategy parameter set of the time slot is generated according to final task allocation of each device and final throughput target, and device running state is monitored, when any device exceeds a preset hard threshold, the strategy parameter set is updated and alarm information is generated; the strategy parameter set comprises: final throughput target, final task allocation, operable load range and time slot control parameter recommended value;

[0010] Step 6, energy consumption settlement and management output is generated based on actual energy consumption of each device and energy consumption budget, including energy consumption cost allocation result, budget execution report and energy saving performance report.

[0011] The application provides a multi-machine linkage device integrated energy-saving management and intelligent decision system, comprising:

[0012] An energy efficiency modeling module is used to obtain running parameters of each device, calculate current unit energy consumption, determine the operable load range of each device, and split the to-be-processed task into standard task units;

[0013] A group-level budget module is used to obtain the group-level residual available energy consumption of the time slot according to the group-level residual available energy consumption of the last time slot, actual energy consumption of each device of the last time slot and group-level newly available energy consumption of the time slot, and establish group-level output constraint and group-level energy consumption constraint;

[0014] An throughput planning module is used to generate initial throughput according to current unit energy consumption, and carry out iterative adjustment in combination with group-level output constraint, group-level energy consumption constraint and set device energy consumption coefficient, to obtain final throughput targets of each device;

[0015] A task allocation module is used to calculate marginal energy consumption cost based on final throughput targets of each device and device energy consumption coefficient, sort devices and allocate standard task units, to obtain final task allocation of each device;

[0016] A strategy alarm module is used to generate strategy parameter set of the time slot according to final task allocation of each device and final throughput target, and monitor device running state, when any device exceeds a preset hard threshold, the strategy parameter set is updated and alarm information is generated; the strategy parameter set comprises: final throughput target, final task allocation, operable load range and time slot control parameter recommended value;

[0017] A settlement report module is configured to generate energy consumption settlement and management outputs based on actual energy consumption and energy consumption budget of each device, including energy consumption cost allocation results, budget execution reports and energy saving performance reports.

[0018] The present application has the advantages that: the present application takes time slots as management granularity, models unit energy consumption, operable load range and standard task unit uniformly, forms final throughput target and final task allocation under the constraint of group-level remaining available energy consumption, and realizes the coordination of resource allocation and output of multiple devices under the same energy consumption budget caliber. Through generating a strategy parameter set containing final throughput target, final task allocation, load range and control parameter suggestion value, and combining hard threshold overrun identification, load range tightening and subset recalculation mechanism, the traceability and stability of strategy adjustment in abnormal scenarios are improved. Further, based on actual energy consumption and energy consumption budget, cost allocation, budget execution and energy saving performance reports are output, forming a closed-loop management link of budget, execution, accounting and evaluation, so that budget deviation, cost responsibility boundary and performance evaluation have a consistent caliber that can be recalculated, thereby improving the refinement and decision-making efficiency of enterprise-level energy consumption management. BRIEF DESCRIPTION OF DRAWINGS

[0019] Figure 1 is a flowchart of the multi-machine linked device integrated energy-saving management and intelligent decision-making method of the present application. DETAILED DESCRIPTION

[0020] The subject matter described herein will now be discussed with reference to example implementations. It should be understood that the discussion of these implementations is merely meant to provide a better understanding of the subject matter described herein and can be changed in function and arrangement without departing from the scope of the present specification. Various processes or components can be omitted, substituted, or added according to various examples. Additionally, features described with respect to some examples can be combined in other examples.

[0021] As shown in Figure 1 the multi-machine linked device integrated energy-saving management and intelligent decision-making method includes the following steps:

[0022] Step 1, obtain the operating parameters of each device, calculate the current unit energy consumption, determine the operable load range of each device, and split the to-be-processed task into standard task units;

[0023] Step 2, obtain the group-level remaining available energy consumption of the current time slot according to the group-level remaining available energy consumption of the previous time slot, the actual energy consumption of each device of the previous time slot, and the group-level newly available energy consumption of the current time slot, and establish group-level output constraints and group-level energy consumption constraints;

[0024] Step 3, generating initial throughput according to current unit energy consumption, and iteratively adjusting in combination with group-level output constraint, group-level energy consumption constraint and set device energy consumption coefficient to obtain final throughput target of each device;

[0025] Step 4, calculating marginal energy consumption cost based on final throughput target of each device and device energy consumption coefficient, sorting and allocating standard task units to devices to obtain final task allocation amount of each device;

[0026] Step 5, generating strategy parameter set of the time slot according to final task allocation amount and final throughput target of each device, and monitoring device running state, when any device exceeds the preset hard threshold, updating the strategy parameter set and generating alarm information; the strategy parameter set includes: final throughput target, final task allocation amount, runnable load interval and time slot control parameter recommended value;

[0027] Step 6, generating energy consumption settlement and management output based on actual energy consumption and energy consumption budget of each device, including energy consumption cost allocation result, budget execution report and energy saving performance report.

[0028] In an embodiment of the present application, the current unit energy consumption is calculated, the runnable load interval of each device is determined, and the to-be-processed tasks are split into standard task units, including:

[0029] Step 11, obtaining running parameters of each device, the running parameters including instant power, instant throughput and to-be-processed task queue length, and calculating current unit energy consumption by the ratio of instant power to instant throughput, when instant throughput is zero, setting current unit energy consumption as a preset upper limit of energy consumption; wherein, instant power is used to reflect the current energy consumption rate of the device, instant throughput is used to reflect the output capacity of the device in the current time slot, and to-be-processed task queue length is used to reflect the total amount of tasks to be completed by the device in the current time slot. Unit energy consumption is used to reflect the energy consumption required per unit output, which is used to measure the energy efficiency level of the device; when the instant throughput of a device is zero in the current time slot, the ratio cannot be directly calculated, and the current unit energy consumption is set as a preset upper limit of energy consumption;

[0030] Step 12, according to the minimum power, maximum power, maximum throughput and adjacent time slot allowed running change amplitude calibrated offline, substituting instant throughput into the adjacent time slot allowed running change amplitude to form the value range of current time slot throughput, and through intersection operation, filtering the value set satisfying the minimum power and maximum power constraints to form the runnable load interval of the current time slot; the runnable load interval is used to limit the throughput range that can be executed by the device in the current time slot, so as to ensure that the device runs in the physically allowed interval;

[0031] Step 13, the length of the to-be-processed task queue is divided by the standard task unit size to obtain the number of standard task units. The standard task unit is a unified task allocation granularity, so that different devices can allocate tasks in a unified scale in subsequent steps, thereby improving the collaborative scheduling efficiency among multiple devices.

[0032] The calculation of the device unit energy consumption, the construction of the device operable load interval, and the standardization processing of the to-be-processed task of the embodiment make the multi-machine system still be able to collaboratively operate under a unified data model under different power levels, different load states, and different margin conditions, thereby providing accurate, controllable, and comparable conditions for subsequent energy-saving decisions.

[0033] In an embodiment of the present application, the group-level residual available energy consumption of the current time slot is obtained according to the group-level residual available energy consumption of the last time slot, the actual energy consumption of each device of the last time slot, and the group-level newly available energy consumption of the current time slot, and the group-level output constraint and the group-level energy consumption constraint are established, including:

[0034] Step 21, the actual energy consumption of each device of the last time slot is summed to reflect the total amount of energy consumption consumed by the group level in the last time slot, and the sum result is deducted from the group-level residual available energy consumption of the last time slot to obtain the initial energy balance without considering the newly added energy consumption amount of the current time slot, and then the group-level newly available energy consumption of the current time slot is added to obtain the group-level residual available energy consumption of the current time slot; the group-level residual available energy consumption is used to represent the total energy consumption amount that can be utilized by the entire multi-machine system in the current time slot;

[0035] Step 22, the throughputs of the devices in the current time slot are summed, and the sum result is limited to be not less than the real-time task demand amount of the current time slot, so as to require the collaborative output of the multiple devices in the current time slot to be sufficient, avoiding task backlog or delayed execution, and establishing the group-level output constraint of the current time slot to ensure that the task execution amount of the current time slot matches the available energy consumption; the real-time task demand amount is used to represent the minimum task amount that must be completed by the system in the current time slot;

[0036] Step 23, the throughputs of the devices in the current time slot are substituted into the corresponding relationship between the throughput and the energy consumption obtained by offline calibration to determine the energy consumption values of the devices in the current time slot, and the sum of all energy consumption values is not greater than the group-level residual available energy consumption of the current time slot, and the group-level energy consumption constraint of the current time slot is established. The corresponding relationship represents the energy consumption level corresponding to different throughputs of the device, which is obtained in advance by testing or modeling; this step ensures that the total energy consumption of the collaborative operation of all devices does not exceed the upper limit of the energy consumption allowed in the current time slot, thereby avoiding the situation of group-level energy consumption overload.

[0037] This embodiment achieves a coordinated balance among various operational objectives in a multi-machine linkage scenario by constructing group-level output constraints and group-level energy consumption constraints. This enables the system to strictly adhere to the energy consumption budget while meeting task execution requirements, thereby improving the energy-saving efficiency and operational stability of multi-device integrated scheduling.

[0038] In one embodiment of the present invention, generating the initial throughput of each device and obtaining the final throughput target of each device includes:

[0039] Step 31: Use the reciprocal of the current unit energy consumption of each device as the allocation weight of each device to reflect the advantage of the device with lower unit energy consumption in terms of energy efficiency, and limit the allocation weight within the preset weight upper limit to avoid abnormally amplified allocation weight of a certain device due to extremely low unit energy consumption. Multiply the ratio of the allocation weight to the total weight by the real-time task demand of this time slot to obtain the initial throughput of each device, and limit the initial throughput within the maximum throughput range of offline calibration, so as to keep the task volume planning consistent with the actual capacity of the device.

[0040] Step 32: Based on the group-level output constraints and group-level energy consumption constraints, determine the energy consumption value of each device by multiplying the device energy consumption coefficient by the initial throughput, and sum all energy consumption values. The device energy consumption coefficient is used to describe the energy consumption level of the device under unit throughput. When the summation result exceeds the group-level remaining available energy consumption in the current time slot, compress the throughput proportionally according to the ratio of each device's energy consumption value to the total energy consumption value of all devices, and limit the compressed throughput within the corresponding operable load range. Repeat the summation and compression process until the group-level output constraints and group-level energy consumption constraints are met. The formula for calculating the compressed throughput is: , This represents the throughput after compression. This represents the throughput before compression. This represents the energy consumption value of all equipment. This indicates the remaining available energy consumption at the group level within this time slot. This indicates that energy consumption has exceeded the limit, which is the difference between the total energy consumption of all devices and the remaining available energy at the group level in this time slot. The compressed throughput will be further limited within its respective operable load range to ensure that the adjusted throughput still meets the physical capacity constraints of the devices. The system repeats the above energy consumption summation and compression process until the throughput of all devices meets the group-level output constraints and group-level energy consumption constraints, thereby achieving a dynamic balance between energy consumption and task output at the group level.

[0041] Step 33: When iteratively adjusting to meet the group-level output constraints and group-level energy consumption constraints, the throughput after iteration is determined as the final throughput target of each device.

[0042] The embodiment realizes data-driven priority sorting and resource allocation optimization of system tasks in the scene of differentiated multi-device running conditions by constructing task allocation weight based on device energy efficiency indicators and generating initial throughput. By setting group-level task output constraints and group-level energy consumption budget constraints, the initial throughput is iteratively corrected, so that the task scheduling meets the overall workload demand while maintaining compliance with the energy consumption budget, thereby realizing collaborative resource management across devices. Further, by gradually compressing and correcting the throughput and combining the consistency verification of the runnable load interval, the system can determine the final task processing target for each device, realize task scheduling optimization and energy consumption resource optimization configuration in a multi-device environment, and build an intelligent scheduling mechanism for business execution efficiency and resource utilization.

[0043] In an embodiment of the present application, the marginal energy consumption cost is calculated, the devices are sorted and the standard task units are allocated to obtain the final task allocation amount of each device, including:

[0044] Step 41, the marginal energy consumption cost is determined according to the product of the device energy consumption coefficient and the final throughput target; the marginal energy consumption cost is used to describe the energy consumption increment brought by increasing unit throughput of the device at the current throughput level;

[0045] Step 42, the devices are sorted in ascending order according to the marginal energy consumption cost, and when the marginal energy consumption costs of two or more devices are the same, the devices are sorted in ascending order according to the device number to form a device sequence, which is used as the standard task unit allocation sequence; the device sequence is used to determine the allocation sequence of the standard task unit among multiple devices, so that the device with lower energy consumption cost can preferentially undertake the task load;

[0046] Step 43, the real-time task demand is segmented according to the standard task unit according to the device sequence, and the residual allocable throughput of the device is checked whether it is lower than the final throughput target unit by unit; when the condition is met, the standard task unit is allocated in sequence until the task unit allocation is completed, and the final task allocation amount of each device is obtained. The standard task unit is the basic unit of task scheduling, which can ensure that the task can be uniformly split among different devices; the residual allocable throughput is used to represent the additional task space that the device can undertake without breaking the physical runnable load interval constraint;

[0047] The embodiment introduces a task allocation strategy based on marginal energy consumption cost to quantitatively evaluate the resource use efficiency of multiple devices in the task execution process, so that the devices with better cost performance can have higher task allocation priority in the scheduling process. Through this strategy, the system can optimize the task load distribution under the overall business constraint condition, so that the multi-device task processing can achieve balance in resource utilization, running cost and business continuity, thereby realizing collaborative task scheduling across devices and improving system running efficiency.

[0048] In one embodiment of the present application, a policy parameter set of the current time slot is generated according to the final task allocation amount of each device and the final throughput target, including:

[0049] In step 51, the final task allocation amount of each device is substituted into the correspondence between the throughput and the control amount obtained from the offline calibration to determine the control parameter suggestion value of each device, and the control parameter suggestion value is limited within the control amount range allowed by the offline calibration of the device.

[0050] Wherein, the final task allocation amount is used to represent the task scale allocated to each device in the current time slot, the correspondence between the throughput and the control amount is used to describe the throughput capacity that the device can achieve under different control amount configurations, and the control parameter suggestion value is used to represent the recommended running parameter configuration of the device in the current time slot under the condition of the current task allocation and the throughput target, which is limited within the control amount range allowed by the offline calibration, so as to ensure that the generated policy parameters have feasibility within the device capability boundary, thereby avoiding the generation of unexecutable or out-of-device constraint configuration suggestions in the policy generation process.

[0051] In step 52, the control parameter suggestion value recorded in the policy parameter set of the previous time slot and the control parameter suggestion value of the current time slot are input, and the maximum allowed variation amplitude of the control amount is determined according to the preset climb-up limit; when the difference between the control parameter suggestion value of the current time slot and the control parameter suggestion value of the previous time slot exceeds the climb-up limit, the transition control parameter suggestion value is set to the control parameter suggestion value of the previous time slot plus the climb-up limit in the deviation direction, otherwise the control parameter suggestion value of the current time slot is taken as the transition control parameter suggestion value.

[0052] Wherein, the climb-up limit is used to limit the maximum variation amplitude of the control parameter suggestion value between adjacent time slots, so as to constrain the adjustment rate of the policy parameter in the time dimension. By taking the control parameter suggestion value of the previous time slot as the reference and introducing the transition control parameter suggestion value, the sudden change of the policy parameter between consecutive time slots can be avoided, so as to ensure the time continuity and comparability of the policy parameter in the budget execution, energy accounting and performance statistics process.

[0053] In step 53, the transition control parameter suggestion value and the control parameter suggestion value of the previous time slot are input, and it is judged whether to update the control parameter suggestion value according to the hysteresis threshold; when the absolute difference between the transition control parameter suggestion value and the control parameter suggestion value of the previous time slot is less than the hysteresis threshold, the control parameter suggestion value of the previous time slot is taken as the control parameter suggestion value of the current time slot, otherwise the transition control parameter suggestion value is taken as the control parameter suggestion value of the current time slot.

[0054] The hysteresis threshold is used to suppress frequent updating of the strategy parameter caused by slight fluctuation of the task allocation amount or the throughput target. When the variation range of the control parameter suggestion value does not exceed the hysteresis threshold, the control parameter suggestion value of the last time slot is kept unchanged, which helps to reduce the jitter of the strategy parameter between adjacent time slots and improve the stability of the strategy parameter set in the management of output and settlement analysis.

[0055] In step 54, the final throughput target, the final task allocation amount, the executable load range and the current time slot control parameter suggestion value are written and formed into the strategy parameter set of the current time slot. The strategy parameter set is used as a unified strategy carrier for the coordinated operation of the multiple devices in the current time slot, and is used to record the task target, load constraint and parameter configuration suggestion of each device in the current time slot, and is used as a direct input data structure for subsequent out-of-limit monitoring, alarm generation, energy consumption budget execution and settlement report generation.

[0056] Through the above steps, the task allocation result, the throughput planning result and the control parameter suggestion can be uniformly organized without changing the physical capability boundary of the device, and the strategy parameter set with time continuity, parameter stability and traceability is formed. The multiple devices can be updated and energy consumption managed under the unified data model and management caliber under different load states and task demand conditions, which provides consistent and calculable data basis for subsequent energy consumption budget execution, cost allocation and energy saving performance report generation.

[0057] In an embodiment of the present application, the device running state is monitored, and when any device exceeds the preset hard threshold, the strategy parameter set is updated and alarm information is generated, including:

[0058] In step 61, the running state vector of each device is monitored, and the running state vector is converted into a scalar state quantity in a preset linear weighting manner. When the scalar state quantity exceeds the corresponding preset hard threshold, the device is marked as an out-of-limit device, the device state is an out-of-limit state, and the final throughput target, the final task allocation amount, the executable load range and the control parameter suggestion value corresponding to the device in the strategy parameter set are recorded.

[0059] The running state vector is composed of multiple state parameters capable of representing the safety and stability of the device running, each state parameter is mapped to a single scalar state quantity through a preset linear weighting manner, and is used to uniformly depict the comprehensive running state of the device at the current time slot. The preset hard threshold is used to limit the safety upper limit of the device running state, and when the scalar state quantity exceeds the corresponding hard threshold, it indicates that the device has a risk of exceeding the expected running interval under the current running condition. By recording the final throughput target, the final task allocation amount, the operable load interval and the control parameter suggestion value when the limit is triggered, the strategy background when the limit occurs can be completely retained, and traceable data basis is provided for subsequent strategy adjustment, alarm analysis and management decision.

[0060] In step 62, for the out-of-limit device, the upper limit of the operable load interval is tightened by a linear proportion of the difference between the out-of-limit amount and the preset hard threshold, to form a tightened operable load interval; and a device subset is constructed with the out-of-limit device and the non-out-of-limit devices that still have residual allocatable throughput within the range of the tightened operable load interval and the original operable load interval.

[0061] The out-of-limit amount is used to represent the degree of the device running state exceeding the preset hard threshold, and the difference between the out-of-limit amount and the hard threshold is mapped to the upper limit of the operable load interval by a linear proportion, to realize quantitative tightening of the load capacity of the out-of-limit device. This method can adaptively adjust the load constraint according to the severity of the out-of-limit, avoiding excessive strategy adjustment caused by fixed proportion or one-time shrinkage. The construction of the device subset is used to limit the scope of subsequent strategy recalculation, so that strategy update is only performed between the out-of-limit device and the related devices with throughput adjustment margin, thereby reducing the influence of global strategy recalculation on the stability of overall task planning and budget execution.

[0062] In step 63, in the device subset, the original operable load interval is replaced by the tightened operable load interval, and the throughput target calculation and standard task unit allocation process are re-executed to obtain the revised final throughput target and final task allocation amount; and the final throughput target, the final task allocation amount and the operable load interval in the strategy parameter set are updated accordingly; wherein the current time slot control parameter suggestion value in the strategy parameter set is determined according to the corresponding relationship between the revised final task allocation amount and the throughput and control amount, and is updated smoothly in combination with the preset climb-up upper limit and the hysteresis threshold, while generating an alarm information.

[0063] The re-execution of the throughput target calculation and task allocation process can rebalance the task allocation relationship among devices under the tightened load constraint condition, so that the system can meet the safety constraint while maintaining the established output target as much as possible. By synchronously updating the throughput target, task allocation amount and load interval in the policy parameter set, it is ensured that the policy parameter set always reflects the current effective management policy state. The smooth updating of the control parameter suggestion value suppresses the violent fluctuation of the policy parameter in the out-of-limit scenario through the climb-up upper limit and hysteresis threshold, thereby improving the continuity and stability of the policy update at the management level. The generated alarm information is used to prompt the upper-layer management system about the out-of-limit event and the corresponding policy adjustment result, so as to support the subsequent analysis and decision of the operation and maintenance personnel or the management system.

[0064] Through the above steps, the running state monitoring, out-of-limit identification, load constraint adjustment and policy parameter set updating can be organically combined in the process of multi-device collaborative operation, so that the out-of-limit processing process is reflected in the form of policy adjustment and alarm output at the management level. Therefore, without destroying the original task planning and budget framework, timely response and traceable management of abnormal running state are realized, and a stable and continuous data basis is provided for subsequent energy consumption settlement, budget execution analysis and energy-saving performance evaluation.

[0065] In an embodiment of the present application, in order to realize quantitative management of the energy-saving performance of multiple devices in different running time slots and provide data basis for cost evaluation, energy-saving assessment and management decision for the system, the system further constructs an energy-saving performance index based on the final task allocation amount, energy consumption value, machine-level remaining available energy consumption and out-of-limit state of each device. The energy-saving performance index is used to reflect the task completion and energy-saving effect of the device under the constraint of the energy consumption budget, and is a comprehensive energy performance evaluation index for the management level.

[0066] The system calculates the task completion degree by ratio of the final task allocation amount of each device and the real-time task demand amount of the corresponding time slot, which is used to represent the load bearing level of the device in the task scheduling environment. The energy consumption deviation is constructed in the form of difference by comparing the energy consumption value generated by each device in the current time slot with the preset machine-level target available energy consumption, which is used to reflect the deviation degree of the actual energy consumption of the device relative to the energy-saving target. The task completion degree and the energy consumption deviation are complementary dimensions, which respectively describe the task contribution and energy consumption of the device.

[0067] The system combines the task completion degree and the energy consumption deviation according to preset weights to obtain a basic performance score. When the energy consumption deviation is lower than zero, i.e., the energy consumption of the device is better than the target value, the system performs gain correction on the basic performance score to strengthen the positive influence of energy-saving performance on the overall performance. When the device has an out-of-limit record in the current time slot, the system further constructs an out-of-limit penalty factor according to the out-of-limit amount, the out-of-limit duration and the tightening amplitude of the load range, and performs penalty correction on the basic performance score, so that the out-of-limit behavior is quantified in the energy-saving evaluation. The introduction of the out-of-limit penalty factor helps the management system to identify the operation risk of the device and promote the device to maintain a stable and controllable energy-saving operation state.

[0068] To quantify the energy-saving stability of the device in the cross-time slot operation process, the embodiment takes whether the machine-level remaining available energy consumption is continuously higher than the machine-level target available energy consumption in multiple consecutive time slots as a judgment condition. When the condition is met, the system performs stability correction on the basic performance score according to a preset stability coefficient, and finally forms an energy-saving performance index to reflect the energy consumption stability and energy-saving consistency of the device in the long-period operation.

[0069] Through the above-mentioned embodiments, the present application introduces a management layer-oriented energy-saving performance evaluation mechanism in the multi-machine linkage energy-saving scheduling framework, so that the device operation data and the enterprise energy management system are linked with each other, which can not only reflect the energy efficiency performance of the device in the task scheduling process, but also provide a quantitative basis for energy-saving decision, performance evaluation and operation optimization, thereby improving the management value and intelligent level of the device integrated energy-saving management system.

[0070] In an embodiment of the present application, based on the actual energy consumption and the energy consumption budget of each device, an energy consumption settlement and management output is generated, including energy consumption cost allocation results, budget execution reports and energy-saving performance reports, including:

[0071] In step 71, the actual energy consumption of each device in the current time slot and the corresponding energy consumption budget are obtained, and the final task allocation amount corresponding to each device is read. The actual energy consumption and the energy consumption budget of each device are subjected to validity check, wherein the actual energy consumption is not less than zero and the energy consumption budget is greater than zero. When any check is not satisfied, the corresponding device is marked as a data abnormal device, and a data abnormality identifier is recorded in the energy consumption settlement and management output. When the above-mentioned checks are all satisfied, the corresponding device is marked as a valid device.

[0072] The actual energy consumption is used to reflect the real energy consumption level of the device in the time slot, and the energy consumption budget is used to represent the energy consumption quota allowed to be used by the device in the time slot, which is determined based on the historical energy consumption level of the device, the final task allocation amount and the final throughput target. By performing effectiveness verification on the actual energy consumption and the energy consumption budget, abnormal metering or budget missing can be identified in advance during the settlement and accounting stage, so as to avoid the influence of abnormal data on subsequent budget execution analysis and cost allocation results. The data anomaly identification is recorded as part of the management output, which is used to distinguish and display abnormal devices in subsequent reports.

[0073] In step 72, for each valid device, a budget deviation value and a budget utilization rate are calculated, wherein the budget deviation value is equal to the difference between the actual energy consumption and the energy consumption budget, and the budget utilization rate is equal to the ratio of the actual energy consumption to the energy consumption budget; and a budget execution state identifier is generated according to the budget utilization rate and a preset budget judgment threshold.

[0074] The budget deviation value is used to quantify the excess or surplus of the actual energy consumption of the device relative to the budget quota, and the budget utilization rate is used to represent the use degree of the device to the budget quota. By comparing the continuous budget utilization rate with the preset budget judgment threshold, a discrete budget execution state identifier is generated, so that the budget execution situation can be reflected in the budget execution report in a unified standard, and the devices in the state of exceeding the budget, approaching the budget or saving are quickly identified by the management personnel.

[0075] In step 73, the group-level actual energy consumption is obtained by summing the actual energy consumption of the valid devices in the time slot, and the group-level energy consumption cost is determined by multiplying the preset energy consumption unit price and the group-level actual energy consumption. For each valid device, an energy consumption proportion and a task proportion are calculated, wherein the energy consumption proportion is equal to the ratio of the actual energy consumption of the device to the group-level actual energy consumption, and the task proportion is equal to the ratio of the final task allocation amount of the device to the sum of the final task allocation amounts of all devices. The energy consumption proportion and the task proportion are weighted by a preset allocation weight to obtain a cost allocation coefficient, and the energy consumption cost allocation result of the device is determined by multiplying the cost allocation coefficient and the group-level energy consumption cost.

[0076] The group-level actual energy consumption is used to represent the overall energy consumption level of multiple devices in the time slot, and the group-level energy consumption cost is used to reflect the cost scale of the overall energy consumption in the financial accounting level. The energy consumption proportion is used to reflect the contribution degree of each device in the total energy consumption, and the task proportion is used to reflect the proportion of each device in the task execution. By weighting and fusing the energy consumption proportion and the task proportion, the energy consumption contribution and the task contribution can be considered in the cost allocation process, so that the energy consumption cost allocation result is more fair and explainable in the management level.

[0077] Step 74, the budget deviation value, the budget utilization rate and the budget execution state identifier are summarized with the data exception identifier to form a budget execution report; and the energy consumption budgets of all effective devices in the time slot are summed up to obtain a group-level energy consumption budget, the difference between the group-level actual energy consumption and the group-level energy consumption budget is calculated as a group-level budget deviation value, the ratio of the group-level actual energy consumption to the group-level energy consumption budget is calculated as a group-level budget utilization rate, and the group-level budget deviation value and the group-level budget utilization rate are written into the budget execution report as group-level summary fields.

[0078] The budget execution report contains device-level budget execution information and group-level budget execution summary information, so that the budget execution situation can be uniformly displayed and compared and analyzed at different management levels. By introducing the group-level budget deviation value and the group-level budget utilization rate, the budget execution effect of the multi-device system in the time slot can be evaluated from the overall level, thereby providing a basis for budget adjustment and energy consumption management decision.

[0079] Step 75, the energy-saving performance index of each device is obtained, the energy-saving performance index is associated with the energy consumption cost allocation result and the budget execution report for writing, thereby forming an energy-saving performance report; wherein the association key is the device identifier and the time slot identifier.

[0080] The energy-saving performance index is used to comprehensively represent the energy-saving performance level of the device in the time slot, and by being associated with the energy consumption cost allocation result and the budget execution information, the energy consumption cost, the budget execution state and the energy-saving performance of the device can be presented in the same report, so that the energy-saving performance report has complete management semantics and traceability.

[0081] Through the above steps, the actual energy consumption of the device, the energy consumption budget, the task allocation result and the energy-saving performance evaluation can be uniformly organized and calculated, thereby forming a multi-level and recalculable energy consumption settlement and management output, so that the budget execution, the cost allocation and the performance evaluation in the multi-device collaborative operation scene are completed under a unified data model and management caliber, thereby providing reliable support for enterprise-level energy consumption management and decision.

[0082] The present application provides a multi-machine linkage device integrated energy-saving management and intelligent decision system, comprising:

[0083] An energy efficiency modeling module is configured to obtain the running parameters of each device, calculate the current unit energy consumption, determine the operable load range of each device, and split the to-be-processed task into standard task units.

[0084] A group-level budget module is configured to obtain the group-level residual available energy consumption in the current time slot according to the group-level residual available energy consumption in the previous time slot, the actual energy consumption of each device in the previous time slot and the group-level newly added available energy consumption in the current time slot, and establish a group-level output constraint and a group-level energy consumption constraint.

[0085] a throughput planning module, configured to generate an initial throughput according to a current unit energy consumption, and iteratively adjust the initial throughput in combination with a group-level output constraint, a group-level energy consumption constraint and a set device energy consumption coefficient to obtain a final throughput target of each device;

[0086] a task allocation module, configured to calculate a marginal energy consumption cost based on the final throughput target of each device and the device energy consumption coefficient, sort and allocate standard task units to the devices to obtain a final task allocation amount of each device;

[0087] a strategy alarm module, configured to generate a strategy parameter set of the current time slot according to the final task allocation amount and the final throughput target of each device, and monitor a device running state, and when any device exceeds a preset hard threshold, update the strategy parameter set and generate an alarm information; the strategy parameter set includes the final throughput target, the final task allocation amount, a runnable load interval and a control parameter suggestion value of the current time slot;

[0088] a settlement report module, configured to generate an energy consumption settlement and management output based on an actual energy consumption and an energy consumption budget of each device, including an energy consumption cost allocation result, a budget execution report and an energy saving performance report.

[0089] It should be noted that the interval and the threshold size are set for ease of comparison, and the size of the threshold depends on the number of sample data and the base number set by the person skilled in the art for each group of sample data, as long as it does not affect the proportional relationship between the parameters and the quantized values. And the above formula is a dimensionless calculation of the value, and the formula is obtained by software simulation of a large amount of data to obtain a formula of the nearest real situation, and the preset parameters in the formula are set by the person skilled in the art according to the actual situation.

[0090] The embodiments of the present application are described above, but the present application is not limited to the above specific embodiments, and the above specific embodiments are only illustrative and not limiting, and a person skilled in the art can make many forms under the inspiration of the present embodiment, which all belong to the protection of the present embodiment.

Claims

1. A method for integrated energy-saving management and intelligent decision-making of multi-machine linkage equipment, characterized in that, Includes the following steps: Step 1: Obtain the operating parameters of each device, calculate the current unit energy consumption, determine the operable load range of each device, and break down the tasks to be processed into standard task units; Step 2: Based on the group-level remaining available energy consumption of the previous time slot, the actual energy consumption of each device in the previous time slot, and the newly added available energy consumption of the group in this time slot, obtain the group-level remaining available energy consumption of this time slot, and establish group-level output constraints and group-level energy consumption constraints. Step 3: Generate the initial throughput based on the current unit energy consumption, and iteratively adjust it in combination with group-level output constraints, group-level energy consumption constraints and set equipment energy consumption coefficients to obtain the final throughput target of each device; Step 4: Calculate the marginal energy consumption cost based on the final throughput target and energy consumption coefficient of each device, sort the devices and assign standard task units to obtain the final task allocation for each device. Step 5: Based on the final task allocation and final throughput target of each device, generate the policy parameter set for this time slot, and monitor the device operation status. When any device exceeds the preset hard threshold, update the policy parameter set and generate alarm information. The set of strategy parameters includes: final throughput target, final task allocation, runnable load range, and recommended values ​​for control parameters in this time slot; Step 6: Based on the actual energy consumption and energy consumption budget of each device, generate energy consumption settlement and management outputs, including energy cost allocation results, budget execution reports and energy-saving performance reports; Step 7: Based on the final task allocation, energy consumption value, machine-level remaining available energy consumption, and over-limit status of each device, construct an energy-saving performance index, including: The ratio of the final task allocation to the real-time task demand is used as the task completion rate, and the energy consumption deviation is constructed by the difference between the energy consumption value and the corresponding machine-level target available energy consumption. The task completion rate and energy consumption deviation are weighted and combined according to preset weights to obtain the basic performance score. When the energy consumption deviation is lower than zero, the basic performance score is corrected. When a device has records of exceeding limits, an over-limit penalty factor is constructed by the amount of over-limit, the duration of over-limit, and the tightening of the operable load range, and the basic performance score is corrected. When the remaining available energy consumption of the generator is higher than the target available energy consumption of the generator for multiple consecutive time slots, the basic performance score is corrected for stability using a preset stability coefficient to obtain the final energy-saving performance index.

2. The integrated energy-saving management and intelligent decision-making method for multi-machine linkage equipment according to claim 1, characterized in that, Calculate the current unit energy consumption, determine the operational load range of each device, and break down the tasks to be processed into standard task units, including: Step 11: Obtain the operating parameters of each device. The operating parameters include real-time power, real-time throughput and the length of the task queue to be processed. Calculate the current unit energy consumption based on the ratio of real-time power to real-time throughput. When the real-time throughput is zero, set the current unit energy consumption to the preset upper limit of energy consumption. Step 12: Based on the offline calibrated minimum power, maximum power, maximum throughput and the allowable operating variation range of adjacent time slots, substitute the real-time throughput into the allowable operating variation range of adjacent time slots to form the value range of the current time slot throughput, filter the value set that satisfies the minimum power and maximum power constraints, and form the operable load range of the current time slot. Step 13: Divide the length of the task queue to be processed into the number of standard task units by integer division according to the size of the standard task unit.

3. The integrated energy-saving management and intelligent decision-making method for multi-machine linkage equipment according to claim 1, characterized in that, Based on the group-level remaining available energy consumption of the previous time slot, the actual energy consumption of each device in the previous time slot, and the newly added group-level available energy consumption in this time slot, the group-level remaining available energy consumption of this time slot is obtained, and group-level output constraints and group-level energy consumption constraints are established, including: Step 21: Sum the actual energy consumption of each device in the previous time slot, subtract the group-level remaining available energy consumption of the previous time slot from the summation result, and then add the newly added group-level available energy consumption in this time slot to obtain the group-level remaining available energy consumption in this time slot. Step 22: Sum the throughput of each device in this time slot, and limit the summation result to be no less than the real-time task demand in this time slot, thus establishing a group-level output constraint for this time slot. Step 23: Substitute the throughput of each device in this time slot into the correspondence between throughput and energy consumption obtained from offline calibration to determine the energy consumption value of each device in this time slot, and sum all the energy consumption values ​​so that the summation result is not greater than the group-level remaining available energy consumption in this time slot, thus establishing the group-level energy consumption constraint for this time slot.

4. The integrated energy-saving management and intelligent decision-making method for multi-machine linkage equipment according to claim 1, characterized in that, Generate the initial throughput of each device to obtain the final throughput target for each device, including: Step 31: Use the reciprocal of the current unit energy consumption of each device as the allocation weight of each device, and limit the allocation weight to the preset weight upper limit. Multiply the ratio of the allocation weight to the total weight by the real-time task demand of this time slot to obtain the initial throughput of each device, and limit the initial throughput to the maximum throughput range calibrated offline. Step 32: Based on the group-level output constraints and group-level energy consumption constraints, determine the energy consumption value of each device by multiplying the device energy consumption coefficient by the initial throughput, and sum all the energy consumption values. When the summation result exceeds the group-level remaining available energy consumption in this time slot, compress the throughput proportionally according to the ratio of each device's energy consumption value to the total energy consumption value of all devices, and limit the compressed throughput within the corresponding operable load range. Repeat the summation and compression process until the group-level output constraints and group-level energy consumption constraints are met. Step 33: When iteratively adjusting to meet the group-level output constraints and group-level energy consumption constraints, the throughput after iteration is determined as the final throughput target of each device.

5. The integrated energy-saving management and intelligent decision-making method for multi-machine linkage equipment according to claim 1, characterized in that, Calculate the marginal energy consumption cost, sort the equipment and assign standard task units to obtain the final task allocation for each equipment, including: Step 41: Determine the marginal energy cost based on the product of the equipment energy consumption coefficient and the final throughput target; Step 42: Sort the devices in ascending order based on their marginal energy consumption cost. When two or more devices have the same marginal energy consumption cost, sort them in ascending order by device number to form a device sequence, which is then used as the standard task unit allocation order. Step 43: Divide the real-time task demand into standard task units according to the device sequence, and check whether the remaining allocable throughput of each device is lower than the final throughput target. If the condition is met, allocate standard task units in sequence until the task units are allocated, and obtain the final task allocation of each device.

6. The integrated energy-saving management and intelligent decision-making method for multi-machine linkage equipment according to claim 1, characterized in that, Based on the final task allocation and final throughput target of each device, a set of strategy parameters for this time slot is generated, including: Step 51: Substitute the final task allocation of each device into the correspondence between throughput and control quantity obtained from offline calibration to determine the recommended control parameter values ​​for each device, and limit the recommended control parameter values ​​to the range of control quantity allowed by the offline calibration of the device. Step 52: The recommended control parameter values ​​recorded in the previous time slot strategy parameter set and the recommended control parameter values ​​of the current time slot are used as inputs. The maximum allowable change range of the control quantity is determined according to the preset ramp-up limit. When the difference between the recommended control parameter value of the current time slot and the recommended control parameter value of the previous time slot exceeds the ramp-up limit, the recommended transition control parameter value is set to the recommended control parameter value of the previous time slot plus the ramp-up limit along the deviation direction. Otherwise, the recommended control parameter value of the current time slot is used as the recommended transition control parameter value. Step 53: Using the suggested value of the transition control parameter and the suggested value of the control parameter of the previous time slot as input, determine whether to update the suggested value of the control parameter based on the hysteresis threshold; when the absolute difference between the suggested value of the transition control parameter and the suggested value of the control parameter of the previous time slot is less than the hysteresis threshold, the suggested value of the control parameter of the previous time slot is used as the suggested value of the control parameter of the current time slot; otherwise, the suggested value of the transition control parameter is used as the suggested value of the control parameter of the current time slot. Step 54: Write the final throughput target, final task allocation, runnable load range, and recommended values ​​of the current time slot control parameters to form the strategy parameter set for this time slot.

7. The integrated energy-saving management and intelligent decision-making method for multi-machine linkage equipment according to claim 1, characterized in that, Monitor device operating status; when any device exceeds a preset hard threshold, update the policy parameter set and generate alarm information, including: Step 61: Monitor the operating status of each device to form an operating status vector, and convert the operating status vector into a scalar status quantity according to a preset linear weighting method; when the scalar status quantity exceeds the corresponding preset hard threshold, mark the device as an over-limit device, the device status is over-limit state, and record the device's final throughput target, final task allocation amount, runnable load range, and the recommended values ​​of the control parameters corresponding to the device in the strategy parameter set. Step 62: For over-limit devices, tighten the upper limit of the operable load range by a linear ratio of the difference between the over-limit amount and the preset hard threshold to form a tightened operable load range; and construct a device subset by using over-limit devices and non-over-limit devices that still have remaining allocable throughput within the tightened operable load range and the original operable load range. Step 63: In the device subset, replace the original runnable load range with a tightened runnable load range, and re-execute the throughput target calculation and standard task unit allocation process to obtain the corrected final throughput target and final task allocation amount; and update the final throughput target, final task allocation amount and runnable load range in the strategy parameter set accordingly; wherein, the suggested value of the current time slot control parameter in the strategy parameter set is determined according to the correspondence between the corrected final task allocation amount and throughput and control amount, and is smoothly updated in combination with the preset ramp-up limit and hysteresis threshold, while generating alarm information.

8. The integrated energy-saving management and intelligent decision-making method for multi-machine linkage equipment according to claim 1, characterized in that, Based on the actual energy consumption and energy budget of each device, energy consumption settlement and management outputs are generated, including energy cost allocation results, budget execution reports, and energy-saving performance reports, including: Step 71: Obtain the actual energy consumption and corresponding energy budget of each device in this time slot, and read the final task allocation for each device; perform validity checks on the actual energy consumption and energy budget of each device, where the actual energy consumption is not less than zero and the energy budget is greater than zero; if any check fails, mark the corresponding device as a data anomaly device and record the data anomaly identifier in the energy consumption settlement and management output; if all the above checks are satisfied, mark the corresponding device as a valid device. Step 72: For each valid device, calculate the budget deviation value and budget utilization rate, where the budget deviation value is equal to the difference between the actual energy consumption and the energy consumption budget, and the budget utilization rate is equal to the ratio of the actual energy consumption to the energy consumption budget; and generate a budget execution status identifier based on the budget utilization rate and the preset budget judgment threshold. Step 73: Based on the sum of the actual energy consumption of the effective devices in this time slot, obtain the actual energy consumption at the group level, and determine the group-level energy consumption cost by multiplying the preset energy consumption unit price by the actual energy consumption at the group level; for each effective device, calculate the energy consumption ratio and the task ratio, where the energy consumption ratio is equal to the ratio of the actual energy consumption of the device to the actual energy consumption at the group level, and the task ratio is equal to the ratio of the final task allocation of the device to the sum of the final task allocations of all devices; weight the energy consumption ratio and the task ratio with a preset allocation weight to obtain the cost allocation coefficient; and determine the energy consumption cost allocation result of the device by multiplying the cost allocation coefficient by the group-level energy consumption cost. Step 74: Summarize the budget deviation value, budget utilization rate, and budget execution status identifier with the data anomaly identifier to form a budget execution report; sum the energy consumption budgets of all valid devices in this time slot to obtain the group-level energy consumption budget; calculate the difference between the actual group-level energy consumption and the group-level energy consumption budget as the group-level budget deviation value; calculate the ratio of the actual group-level energy consumption to the group-level energy consumption budget as the group-level budget utilization rate; and write the group-level budget deviation value and the group-level budget utilization rate into the budget execution report as group-level summary fields. Step 75: Obtain the energy-saving performance index of each device, and link the energy-saving performance index with the energy consumption cost allocation result and budget execution report to form an energy-saving performance report; the linking key is the device identifier and the time slot identifier.

9. A multi-machine integrated energy-saving management and intelligent decision-making system, characterized in that, The method for integrated energy-saving management and intelligent decision-making of multi-machine linkage equipment as described in any one of claims 1-8 includes: The energy efficiency modeling module is used to obtain the operating parameters of each device, calculate the current unit energy consumption, determine the operating load range of each device, and break down the tasks to be processed into standard task units. The group-level budget module is used to obtain the remaining available energy consumption of the group in the current time slot based on the remaining available energy consumption of the group in the previous time slot, the actual energy consumption of each device in the previous time slot, and the newly added available energy consumption of the group in the current time slot, and to establish group-level output constraints and group-level energy consumption constraints. The throughput planning module is used to generate the initial throughput based on the current unit energy consumption, and iteratively adjust it by combining the group-level output constraints, group-level energy consumption constraints and the set equipment energy consumption coefficients to obtain the final throughput target of each device. The task allocation module is used to calculate the marginal energy consumption cost based on the final throughput target and energy consumption coefficient of each device, sort the devices and allocate standard task units to obtain the final task allocation amount for each device. The strategy alarm module is used to generate a set of strategy parameters for the current time slot based on the final task allocation and final throughput target of each device, and to monitor the operating status of the devices. When any device exceeds a preset hard threshold, the strategy parameter set is updated and an alarm message is generated. The strategy parameter set includes: final throughput target, final task allocation, runnable load range, and recommended values ​​of control parameters for the current time slot. The settlement report module is used to generate energy consumption settlement and management outputs based on the actual energy consumption and energy consumption budget of each device, including energy cost allocation results, budget execution reports and energy-saving performance reports.

Images