Thermal power plant auxiliary machine group cooperative optimization operation method and system

Abstract

The application provides a thermal power plant auxiliary machine group cooperative optimization operation method and system, and belongs to the thermal power plant operation technical field.The method comprises the following steps: collecting multi-dimensional original data and generating a standardized data set; establishing a multi-physical field coupling model and generating an energy routing model with complete constraints; splitting a plurality of cost sub-items and configuring corresponding weight coefficients to form a quantifiable multi-objective optimization function through weighted fusion; performing differential quantum bit coding and designing hierarchical population initialization and adaptive quantum gate operators; performing parallel global optimization in a super-high-dimensional mixed solution space to generate an optimal scheduling scheme and a plurality of alternative schemes; and grading and issuing standardized control instructions and performing sequential control execution.The application solves the limitations of traditional static modeling and conventional algorithms easily falling into local optimization, achieves the purpose of global optimal scheduling solution, realizes the dynamic balance of multiple requirements of cost reduction, carbon reduction and equipment life extension, and eliminates the hidden dangers of equipment over-limit and pipe network fluctuation.

Description

Technical Field

[0001] This invention relates to the field of thermal power plant operation technology, and more specifically to a method and system for the coordinated and optimized operation of auxiliary equipment groups in thermal power plants. Background Technology

[0002] As the core energy-consuming equipment cluster for power generation, heating, and environmental management, the auxiliary equipment group of a thermal power plant, together with the plant's electricity, heat, cooling, and gas coupled microgrid system, constitutes the main body of the power plant's energy operation. Its coordinated dispatch level directly determines the power plant's energy consumption economy, environmental compliance, and system operation stability. With the advancement of the construction of new power systems, the dynamic fluctuation of grid time-of-use electricity prices, the continuous tightening of environmental carbon emission control, and the frequent switching of unit peak-shaving operating conditions, traditional thermal power plants generally adopt a dispatch mode of independent setpoint control of auxiliary equipment and fragmented operation of multi-energy systems. They lack a comprehensive coupled and coordinated optimization system, making it difficult to adapt to complex dynamic operating scenarios, and their intelligent and refined dispatch capabilities have obvious shortcomings.

[0003] In existing technologies, the operation of auxiliary equipment groups in thermal power plants still suffers from the following shortcomings: First, the ability to accurately model and intelligently optimize is insufficient. Traditional scheduling relies mainly on static modeling of equipment rated parameters, ignoring the nonlinear characteristics of auxiliary equipment groups under varying operating conditions and the coupling relationship of multiple energy flows. Moreover, conventional optimization algorithms are difficult to adapt to high-dimensional and complex scheduling scenarios, easily getting trapped in local optima and having low convergence accuracy, making it impossible to achieve a globally optimal scheduling solution. Second, the optimization and control dimensions are singular. Existing technologies often unilaterally pursue the reduction of plant power consumption, without establishing a dynamic balance mechanism for multiple objectives such as economy, environmental protection, and safety. They cannot adaptively adjust scheduling strategies according to scenarios such as electricity price fluctuations, heavy pollution warnings, and grid supply guarantees, making it difficult to simultaneously meet the multiple needs of cost reduction, carbon reduction, and equipment life extension. Third, the system's adaptability and safety implementation capabilities are weak. Traditional scheduling model parameters are fixed and lack closed-loop iterative update capabilities, making it unable to adapt to long-term changes such as equipment aging and changing operating conditions. The scheduling accuracy decreases year by year, and there is a lack of a sound safety verification and smooth execution mechanism. The implementation of optimization schemes is prone to causing safety hazards such as equipment exceeding limits and pipeline fluctuations, resulting in poor engineering practicality. Summary of the Invention

[0004] The purpose of this invention is to provide a method and system for the collaborative optimization operation of auxiliary equipment groups in thermal power plants, in order to solve the technical problems that existing static modeling and traditional algorithms cannot achieve global optimization in multi-energy coupled scenarios, lack a multi-objective dynamic trade-off control mechanism, and lack safety control over the implementation of optimization instructions.

[0005] To achieve the above objectives, this invention provides a method for the collaborative optimization operation of auxiliary equipment groups in thermal power plants. The method includes: collecting multi-dimensional raw data using a global sensor network deployed across the target thermal power plant's auxiliary equipment group, and preprocessing the data through edge computing to generate a standardized dataset; based on the standardized dataset and combined with the factory performance parameters of the target thermal power plant's auxiliary equipment group, establishing a multi-physics coupling model for each auxiliary unit, and constructing a multi-layered constraint equation set to generate an energy routing model with complete constraints; based on the energy routing model and the standardized dataset, decomposing multiple cost components and configuring corresponding weight coefficients, and forming a quantifiable and solvable multi-objective optimization function through weighted fusion. To address the high nonlinearity and multi-objective optimization characteristics of the energy routing model, the objective decision variables are encoded using differentiated qubits based on the principle of quantum superposition. A hierarchical population initialization and adaptive quantum gate operator are designed, and an observation decoding and hierarchical fitness evaluation process is established. Based on the adaptive quantum gate operator, combined with the energy routing model and multi-objective optimization function, parallel global optimization is performed in the ultra-high-dimensional hybrid solution space. A preset process is executed iteratively to generate the optimal scheduling scheme and multiple alternative schemes. The optimal scheduling scheme and multiple alternative schemes are parsed into standardized control commands, subjected to multiple security checks, and the standardized control commands are distributed hierarchically and executed sequentially, with execution status feedback and hierarchical anomaly handling.

[0006] Optionally, the global sensor network adopts a three-layer distributed three-dimensional deployment architecture, including an auxiliary equipment body perception layer, a plant microgrid topology perception layer, and an environment and external market perception layer. Based on the global sensor network deployed in the target thermal power plant's auxiliary equipment group, multi-dimensional raw data is collected, including: the deployment of measurement points for the target energy-consuming auxiliary equipment in the target thermal power plant's auxiliary equipment group, and the division of monitoring units according to three major monitoring dimensions: electrical, mechanical, and fluid, serving as the auxiliary equipment body perception layer; the deployment of data collection points for the overall power grid structure and multi-energy flow pipeline topology of the target thermal power plant, divided into power supply side, heating and steam network side, and energy storage side, constructing multi-energy production, transmission, and distribution links within the target thermal power plant area, serving as the plant microgrid topology perception layer and the environment layer; and the access of external variable data from the natural environment and policy / market dimensions of the target thermal power plant area, eliminating the interference of environmental and policy changes on the optimization results, serving as the environment and external market perception layer.

[0007] Optionally, the step of establishing a multi-physics coupling model for each target thermal power plant auxiliary machine, constructing a multi-layer constraint equation set, and generating an energy routing model with complete constraints includes: based on directed weighted graph theory, performing microgrid topology decomposition, decomposing it into multiple types of topology node sets and multiple types of energy transmission directed edge sets, with each node and transmission directed edge attached with corresponding attribute parameters to define the energy routing network topology; based on the energy routing network topology, digitally integrating full performance curves, drift mapping variable operating condition characteristics, and defining two-dimensional state variables to establish a multi-physics coupling model corresponding to each target thermal power plant auxiliary machine; based on the multi-physics coupling model, constructing a multi-layer constraint equation set from four dimensions: energy conservation, equipment limits, pipeline transmission limits, and plant-wide supply and demand balance, generating an energy routing model with complete constraints, so that all scheduling schemes derived from the model meet the on-site physical operation criteria.

[0008] Optionally, the multiple cost items include an economic objective sub-function, an environmental objective sub-function, and a technical safety objective sub-function. Based on the energy routing model and standardized dataset, multiple cost items are separated and corresponding weight coefficients are configured. A quantifiable multi-objective optimization function is formed through weighted fusion. This includes: the economic objective sub-function, which is the sum of direct production costs and implicit equipment loss costs throughout the entire lifecycle, divided into three independent calculation modules: time-of-use plant power costs, auxiliary equipment start-up and shutdown loss costs, and variable operating condition related maintenance costs; the environmental objective sub-function is constructed through three dimensions: carbon emission control, air pollutant compliance, and implicit costs of solid waste disposal, with all parameters dynamically priced based on carbon trading conditions and environmental control strategies; the technical safety objective sub-function uses tiered penalty coefficients to quantify risk losses, setting multi-level penalty items from four dimensions: main unit operation assurance, single auxiliary equipment health, plant power grid power quality, and overall plant heating network hydraulic stability; and corresponding weight coefficients are configured for the economic, environmental, and technical safety objective sub-functions, forming a quantifiable multi-objective optimization function through weighted fusion.

[0009] Optionally, the weighting coefficient is adaptively and dynamically adjusted through multi-dimensional triggering signals of the auxiliary equipment group of the thermal power plant, and multi-level trigger preset thresholds and weight adjustment gradients are set to divide multiple control scenarios. The control scenarios include peak electricity price control period of the power grid, heavy pollution weather graded early warning scenario, peak summer and winter heating supply guarantee period, carbon price rise, unit planned maintenance window period and environmental emission monthly assessment period.

[0010] Optionally, the step of using the quantum superposition principle to encode the target decision variables in differentiated qubits and designing hierarchical population initialization and adaptive quantum gate operators includes: using a classification and block-based qubit encoding method to map the target decision variables in the energy routing model to the corresponding qubit probability amplitude space and dividing it into multiple encoding blocks; using a hierarchical differentiated superposition state initialization strategy to distinguish the differentiated initial values ​​of multiple encoding blocks, so as to take into account both the common sense of the target thermal power plant equipment operation and the breadth of the global search; using the quantum rotation gate as the global master evolution operator and the quantum NOT gate as the auxiliary mutation operator, and combining the convergence law of the target thermal power optimization to design an adaptive parameter adjustment rule, so that the operator parameters change dynamically with the iteration number and the population diversity index.

[0011] Optionally, the parallel global optimization in the ultra-high-dimensional hybrid solution space, cyclically executing a preset process to generate an optimal scheduling scheme and multiple alternative schemes, includes: an adaptive quantum gate operator performing a coarse global exploration in the early stage and a fine local mining in the later stage; a large-scale traversal of the ultra-high-dimensional hybrid solution space to select high-quality regions in the early stage and fine-tuning parameters of high-quality regions in the later stage; batch generating an initial quantum population according to the rules of hierarchical population initialization, with the number of the initial quantum population adaptively adjusted according to the dimension of the optimization variable; using a multi-threaded concurrent mode, performing quantum measurement collapse and binary decoding on all quantum individuals in the initial quantum population to split and obtain a scheduling scheme; sorting the target population according to the fitness value in a preset order, selecting the best quantum individuals in the target population, and comparing it with the historical best scheme; calculating the target parameters based on the fitness of the best quantum individuals and the overall dispersion data of the population, and batch generating offspring populations; continuously executing cyclic iterations, stopping the operation when any preset termination condition is met, and obtaining the optimal scheduling scheme and multiple alternative schemes.

[0012] Optionally, the step of parsing the optimal scheduling scheme and multiple alternative schemes into standardized control instructions and performing multiple security checks includes: based on the target thermal power plant auxiliary equipment group database and the auxiliary equipment control hardware type, decomposing the optimal scheduling scheme into instructions and generating control instructions differentiated by various types of equipment; adding various tags to all control instructions, distinguishing between immediate execution instructions and timed delayed execution instructions to obtain standardized control instructions; and sending the standardized control instructions to a preset security check engine for multiple security checks. If any check fails, an alarm is issued and an anomaly log is retained.

[0013] Optionally, the collaborative optimization operation method further includes: collecting the operation data of the target thermal power plant auxiliary equipment group after executing standardized control commands based on the deployed global sensor network; using digital twin technology to compare the deviation between the predicted value and the operation data; automatically correcting the energy routing model parameters and adaptive quantum gate operator parameters; and realizing closed-loop self-optimization of the entire system.

[0014] On the other hand, the present invention provides a collaborative optimization operation system for auxiliary equipment groups of thermal power plants. The collaborative optimization operation system includes a control module, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor. The processor executes the computer program to implement the collaborative optimization operation method for auxiliary equipment groups of thermal power plants as described above.

[0015] Through the above technical solutions, this invention collects multi-dimensional operating condition data through a global sensor network, constructs a multi-physics coupling model and an energy routing model, and combines the quantum superposition principle and adaptive quantum gate operators to achieve parallel global optimization in ultra-high-dimensional hybrid solution space. This solves the limitations of traditional static modeling and conventional algorithms that are prone to getting trapped in local optima, achieving the goal of globally optimal scheduling and solution. Through multi-objective optimization functions, it adapts to scenarios such as electricity price fluctuations, heavy pollution early warning, and power grid supply security, achieving a dynamic balance between multiple demands for cost reduction, carbon reduction, and equipment life extension. By comparing prediction and execution deviations through digital twin technology, it automatically corrects model and algorithm parameters, solving the problem of accuracy decay caused by model solidification. Through standardized instruction parsing, multiple security checks, and sequential control execution, it eliminates the hidden dangers of equipment exceeding limits and pipeline fluctuations, significantly improving engineering practicality.

[0016] Other features and advantages of the present invention will be described in detail in the following detailed description section. Attached Figure Description

[0017] The accompanying drawings are provided to further illustrate embodiments of the present invention and form part of the specification. They are used together with the following detailed description to explain the embodiments of the present invention, but do not constitute a limitation thereof. In the drawings: Figure 1 This is a flowchart illustrating the method for coordinated and optimized operation of auxiliary equipment groups in thermal power plants according to the present invention. Figure 1 ; Figure 2 This is a schematic diagram of the process for collecting multi-dimensional raw data in this invention; Figure 3 This is a schematic diagram of the process for generating the energy routing model in this invention; Figure 4 This is a schematic diagram of the process for generating multi-objective optimization functions in this invention; Figure 5 This is a schematic diagram of the process for designing adaptive quantum gate operators in this invention; Figure 6 This is a schematic diagram of the process for generating the optimal scheduling scheme and multiple alternative schemes in this invention; Figure 7 This is a flowchart illustrating the process of parsing standardized control commands and performing multiple security checks in this invention. Figure 8This is a flowchart illustrating the method for coordinated and optimized operation of auxiliary equipment groups in thermal power plants according to the present invention. Figure 2 . Detailed Implementation

[0018] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit the scope of the present invention.

[0019] It should be noted that the acquisition, transmission, storage, use, and processing of data in the technical solution of this application all comply with the relevant provisions of national laws and regulations. In the embodiments of this application, certain existing industry solutions such as software, components, and models may be mentioned. These should be considered exemplary, intended only to illustrate the feasibility of implementing the technical solution of this application, and do not imply that the applicant has already used or necessarily used such solutions.

[0020] Please refer to Figure 1 and Figure 8 This invention provides a method for the coordinated and optimized operation of auxiliary equipment groups in thermal power plants. The method may include: Step S100: Based on the global sensor network deployed in the auxiliary equipment group of the target thermal power plant, collect multi-dimensional raw data, and perform data preprocessing through edge computing to generate a standardized dataset.

[0021] Please refer to Figure 2 In this embodiment of the invention, the global sensor network adopts a three-layer distributed three-dimensional deployment architecture, which may include an auxiliary machine body perception layer, a plant microgrid topology perception layer, and an environment and external market perception layer. Based on the global sensor network deployed in the auxiliary machine group of the target thermal power plant, multi-dimensional raw data is collected, which may include: Step S110: Set up measurement points for the target energy-consuming auxiliary equipment of the target thermal power plant auxiliary equipment group, and divide the monitoring units according to the three major monitoring dimensions of electrical, mechanical and fluid, as the sensing layer of the auxiliary equipment body.

[0022] In a preferred embodiment of the present invention, the target energy-consuming auxiliary equipment may include a forced draft fan, an induced draft fan, a primary air fan, a steam-driven / electric feedwater pump, a condensate pump, a circulating water pump, a desulfurization booster fan, etc. The auxiliary equipment body sensing layer may include an electrical monitoring unit, a mechanical condition monitoring unit, and a fluid parameter monitoring unit. For the electrical monitoring unit, a high-precision 0.2S-level three-phase smart energy meter and an online harmonic monitoring device can be installed at the 6kV / 380V auxiliary equipment drive motor incoming switch cabinet and terminal box to continuously collect parameters such as three-phase line voltage, operating current of each phase, active power, reactive power, and real-time power factor.

[0023] In a preferred embodiment of the present invention, for the mechanical condition monitoring unit, piezoelectric accelerometers, magnetoelectric velocity sensors, and PT100 platinum resistance temperature sensors can be respectively installed at key measuring points on the auxiliary machine drive / non-drive end bearing housing, motor housing, fan casing, and pump base. This allows for the continuous acquisition of characteristic parameters such as bearing horizontal / vertical / axial three-dimensional vibration intensity, full-frequency vibration spectrum, bearing outer / inner ring operating temperature, motor stator and rotor winding temperature, and gearbox oil temperature.

[0024] In a preferred embodiment of the present invention, for the fluid parameter monitoring unit, high-precision differential pressure transmitters and electromagnetic flowmeters can be installed at key points in the auxiliary machine inlet and outlet main pipes, bypass pipes, and minimum flow circulation pipes to collect fluid parameters such as medium inlet and outlet pressures, pipeline pressure differentials, instantaneous volume / mass flow rates of the medium, working fluid temperature, and container liquid level in real time.

[0025] Step S120: For the overall power grid and multi-energy flow pipeline topology of the target thermal power plant, data acquisition points are deployed on the power supply side, the heating and steam network side, and the energy storage side to construct the multi-energy production, transmission, and distribution links within the target thermal power plant area, serving as the plant area microgrid topology sensing layer and environment.

[0026] In a preferred embodiment of the present invention, during power supply side monitoring, the three-phase voltage, system frequency, power angle, and active / reactive power flow data of the main transformer high-voltage side, high-voltage plant service bus, and low-voltage plant service segment bus can be collected in real time; the system can also connect to the power grid dispatching system in real time to capture regional time-of-use electricity prices, AGC unit load adjustment instructions, AGC bus voltage control instructions, power grid reserve capacity requirements, and system power deficit early warning signals.

[0027] In a preferred embodiment of the present invention, for the heating steam network side, pressure transmitters and high-precision thermocouples can be installed at the boiler main steam outlet, reheat steam header, turbine extraction steam outlet, plant heating station, and industrial steam supply branch pipelines to continuously collect the medium pressure, temperature, and instantaneous steam flow of main steam, reheat steam, heating extraction steam, and industrial extraction steam, and calculate the total heat production of the entire plant, pipeline heat dissipation loss, and actual heat load consumption.

[0028] In a preferred embodiment of the present invention, during energy storage side monitoring, the system can collect data in real time on distributed energy storage devices in the plant area, such as lithium iron phosphate energy storage battery packs, flywheel energy storage devices, high-temperature hot water storage tanks, and compressed air energy storage tanks. This data includes individual cell voltage, overall state of charge (SOC), battery health (SOH), real-time charging and discharging power, remaining available energy storage time, and tank medium temperature and pressure. This allows for precise control of the energy storage system's adjustability margin and supports energy storage participation in the optimized scheduling of plant power peak shaving and valley filling.

[0029] Step S130: Access external variable data from the two dimensions of the target thermal power plant area, namely the natural environment and policy market, and eliminate the interference of environmental and policy changes on the optimization results, serving as the environmental and external market perception layer.

[0030] In a preferred embodiment of the present invention, the environmental and external market perception layer can be divided into meteorological monitoring and market and policy signals. For meteorological monitoring, integrated micro-weather stations are deployed in open areas of the factory to collect ambient dry-bulb temperature, relative humidity, real-time wind speed and direction, atmospheric pressure, and solar irradiance around the clock. These environmental parameters are used to correct for the environmentally variable operating efficiency of air-cooled units, circulating water systems, and fan equipment. For market and policy signals, a data interface can be used to connect to the national carbon emission trading system to capture real-time carbon quota transaction prices and simultaneously access data from the local environmental protection bureau's environmental management platform to obtain daily pollutant emission limits, heavy pollution weather warning levels, and ultra-low emission assessment indicators.

[0031] In this embodiment of the invention, data preprocessing via edge computing can include data cleaning and repair, time alignment and resampling, standardization and normalization, and feature extraction and construction. During data cleaning and repair, the improved Raida 3σ criterion and quartile box plot anomaly identification algorithm are integrated to accurately distinguish three types of abnormal data: instantaneous sensor disconnection, electromagnetic interference jumps, and constant dead zones caused by probe jamming. During time alignment and resampling, since the sampling cycles of voltage meters, vibration sensors, flow meters, and weather station equipment are inconsistent (e.g., a range from milliseconds to minutes), the edge gateway completes full data resampling according to a unified 5-second basic scheduling cycle. During standardization and normalization, the Min-Max linear mapping standardization algorithm can be used to uniformly map all original feature data to the [0,1] numerical range, eliminating weight bias caused by differences in dimensions and numerical magnitudes. During feature extraction and construction, the instantaneous power consumption of a single auxiliary machine can be calculated based on real-time voltage and current values, and mathematical conversions can be performed on the vibration spectrum to solve for equipment fault characteristic quantities such as kurtosis, impulse factor, and margin coefficient.

[0032] Step S200: Based on the standardized dataset and combined with the factory performance parameters of the auxiliary equipment group of the target thermal power plant, establish a multi-physics coupling model for each auxiliary equipment of the target thermal power plant, and construct a multi-layer constraint equation set to generate an energy routing model with complete constraints.

[0033] Please refer to Figure 3 In this embodiment of the invention, a multiphysics coupling model is established for each target thermal power plant auxiliary machine, a multi-layer constraint equation set is constructed, and an energy routing model with complete constraints is generated, which may include: Step S210: Based on the directed weighted graph theory, perform microgrid topology decomposition, which is divided into multiple sets of topology nodes and multiple sets of directed edges for energy transmission. Each node and directed edge for transmission is accompanied by corresponding attribute parameters to define the energy routing network topology.

[0034] In a preferred embodiment of the present invention, the set of multiple types of topological nodes may include source nodes (e.g., energy supply nodes), sink nodes (e.g., energy consumption terminal nodes), conversion nodes (e.g., energy conversion nodes that convert between different energy forms), and storage nodes (e.g., energy buffer nodes). For the set of directed edges for multiple types of energy transmission, all energy transmission carriers are abstracted as weighted directed transmission edges, which are divided into five categories: power cable branches, high-temperature steam pipelines, heating and hot water pipelines, cold water transmission pipelines, and compressed air pipelines. Each transmission edge is preset with three inherent attributes: the upper limit of the rated maximum transmission capacity of the pipeline / cable, the medium transmission loss coefficient along the path, and the energy transmission lag time.

[0035] Step S220: Based on the energy routing network topology, digitally integrate the full performance curves, drift map the variable operating condition characteristics, define two-dimensional state variables, and establish a multi-physics coupling model corresponding to each target thermal power plant auxiliary machine.

[0036] In a preferred embodiment of the present invention, when digitally integrating the full performance curves, the three characteristic curves of auxiliary equipment's factory-matched flow-head, flow-shaft power, and flow-overall efficiency can be entered into the model database through polynomial fitting and multidimensional lookup table method. The model interpolates and solves the output work and input power consumption of the equipment under the corresponding working conditions in real time according to the adjustment commands such as frequency conversion, inlet baffle opening, and hydraulic coupling opening, so as to accurately reproduce the energy consumption law of the equipment under rated working conditions and variable working conditions.

[0037] In a preferred embodiment of the present invention, when drift mapping changes the operating conditions, the performance degradation law after the equipment deviates from the design operating conditions can be fully considered. For example, nonlinear characteristics such as the sharp drop in efficiency of water pump in the low flow range, the surge in energy consumption of fan in the stall range, and the significant drop in power factor of motor under low load are all considered. The characteristic mapping is completed by piecewise high-order polynomial fitting formula, avoiding the disconnect between the optimization results and the actual operating conditions on site caused by the model relying solely on the rated parameters.

[0038] In a preferred embodiment of the present invention, when defining dual-dimensional state variables, two types of controllable variables can be set for a single auxiliary machine: discrete Boolean state variables (e.g., 0 = standby, 1 = normal operation) and continuous optimization variables (e.g., real-time load rate, motor operating speed, outlet medium pressure, medium transport flow rate), and all variables are used as decision variables for subsequent optimization algorithms.

[0039] Step S230: Based on the multiphysics coupling model, a multi-layer constraint equation set is built from four dimensions: energy conservation, equipment limits, pipeline transmission limits, and plant supply and demand balance. This generates an energy routing model with complete constraints, ensuring that all scheduling schemes derived from the model meet the on-site physical operation criteria.

[0040] In a preferred embodiment of the present invention, for the constraint of node energy conservation, any topological node follows the energy balance equation: total energy input to the node = total energy output of the node + various losses of the node itself (e.g., motor heat loss, mechanical friction loss, pipeline heat dissipation loss). For example, for a water pump conversion node: input electrical energy = output water potential energy + motor copper and iron loss + pump body hydraulic loss, and the energy of the entire node is calculated in a closed loop.

[0041] In a preferred embodiment of the present invention, the physical operation constraints of the equipment may include upper and lower limits of output (e.g., the operating load of a single auxiliary machine is limited to between the minimum safe output and the rated maximum output, such as the minimum operating flow of an electric water pump shall not be less than 30% of the rated flow), load ramp rate constraints (e.g., the maximum adjustment range of the auxiliary machine load per unit time is limited, such as the load change of the water pump per minute is ≤ 10% of the rated power, to prevent sudden load changes from causing large fluctuations in the main pipe pressure and pipeline impact), and start-up and shutdown frequency and time constraints (e.g., setting the maximum number of start-ups and shutdowns of a single auxiliary machine per day, while constraining the minimum continuous operating time and the downtime of the equipment).

[0042] In a preferred embodiment of the present invention, network transmission constraints may include that the real-time transmission power of each cable and pipeline transmission branch does not exceed the upper limit of the rated thermal stability / pressure bearing capacity, and overcapacity solutions are directly determined as infeasible solutions; the overall plant supply and demand balance constraints can be expressed as: total on-grid power generation of the unit = total power generation of the boiler - total real-time power consumption of the plant; total heat production of the plant = total real-time heat load demand of the plant area + heat dissipation loss of the entire pipeline network, ensuring the overall supply and demand balance of the two core energy sources, electricity and heat.

[0043] For example, a simplified energy routing topology model for a 660MW combined heat and power (CHP) unit can be constructed: When dividing the topology nodes, the source node is the external grid power grid + coal-fired boiler; the core conversion node is the electric feedwater pump; the sink nodes are divided into the turbine steam consumption and power generation node and the plant area residential heating consumption node; the transmission edges are 6kV plant auxiliary power cables and high-temperature main steam transmission pipelines. The energy routing logic can be as follows: External grid power is transmitted to the feedwater pump conversion node via the plant auxiliary power cable, and after power consumption, it is converted into high-pressure feedwater potential energy and sent to the turbine to complete the thermodynamic cycle; a portion of the high-temperature steam produced by the boiler enters the turbine to generate electricity, and the remaining steam is de-cooled and depressurized before being transported through the steam pipeline to the heating sink node to supply heating to the plant area. When configuring constraint parameters: the rated shaft power of the feedwater pump is 5000kW, the lower limit of the model operating load is 1500kW and the upper limit is 5000kW, and the maximum load adjustment value per minute is 500kW; the rated maximum steam production of the boiler is 2000t / h, and the upper limit flow rate of the steam transmission pipeline of the heating branch is 65t / h. All constraint parameters are written into the model constraint equation set.

[0044] Step S300: Based on the energy routing model and standardized dataset, multiple cost items are separated and corresponding weight coefficients are configured. Through weighted fusion, a multi-objective optimization function that can be quantified and solved is formed.

[0045] Please refer to Figure 4 In this embodiment of the invention, the multiple cost components may include economic objective sub-functions, environmental objective sub-functions, and technical safety objective sub-functions. Based on the energy routing model and standardized dataset, multiple cost components are decomposed and corresponding weight coefficients are configured. A quantifiable multi-objective optimization function is formed through weighted fusion, which may include: Step S310: The economic objective sub-function is the sum of the direct production cost and the implicit equipment loss cost throughout the entire cycle, which is divided into three independent calculation modules: time-of-use plant power cost, auxiliary machine start-up and shutdown loss cost, and variable operating condition related maintenance cost.

[0046] In a preferred embodiment of the present invention, for the time-of-use plant power cost accounting module, the operating system can connect to the regional power grid marketing system in real time to obtain the settlement electricity price for industrial and commercial electricity in three time periods of the day (e.g., peak, flat, and valley periods). It then breaks down the instantaneous power consumption of all auxiliary plant equipment and calculates the basic electricity purchase cost by accumulating the costs in segments according to the differentiated unit price for different time periods. For flexible and adjustable loads such as desulfurization booster fans and plant circulating water pumps that can temporarily avoid peak electricity demand, an additional load interruption compensation pricing model is configured: if the dispatching scheme selects to shut down during peak hours to avoid peak electricity demand, the saved electricity purchase cost, after deducting the compensation cost for purchasing temporary power sources from external sources, is included in the cost optimization benefit; for auxiliary loads connected to the plant's distributed photovoltaic self-generated electricity, the photovoltaic power cost is used to replace the external grid electricity purchase cost, and the actual electricity expenditure is deducted accordingly.

[0047] In a preferred embodiment of the present invention, regarding the auxiliary machine start-stop loss cost accounting module, the stator windings of large rotating auxiliary machines undergo a hot-cold cycle from room temperature to high temperature and rapid cooling upon startup and shutdown. The shaft system is subjected to starting impact torque, and frequent start-stops accelerate insulation aging and bearing fatigue damage. This application establishes an equivalent depreciation calculation formula for a single start-stop of a single piece of equipment based on the equipment manufacturer's life test data. Insulation loss, bearing wear, and mechanical impact loss are converted into corresponding capital costs. Each start-stop adds a fixed loss cost. When the number of start-stops per day exceeds the limit, a tiered pricing conversion factor is activated. The more frequent the start-stops, the higher the marginal conversion cost. From an economic constraint perspective, this prevents the optimization algorithm from blindly and frequently starting and stopping large auxiliary machines for short-term power saving.

[0048] In a preferred embodiment of the present invention, the variable operating condition associated maintenance cost accounting module, combined with the real-time operating efficiency, vibration intensity, bearing temperature, and other health characteristic parameters of the auxiliary machine collected and extracted from the database, divides the operating conditions into three zones: a high-efficiency economic operating zone, a normal operating zone, and a deterioration and loss zone. When the auxiliary machine deviates from the manufacturer's design high-efficiency operating condition for a long period of time, impeller erosion, pipe fouling, and motor heating rate increase significantly, the equipment maintenance cycle shortens, and the consumption of maintenance consumables increases. The system pre-enters an annual average maintenance quota database under different operating conditions, and linearly interpolates the daily equivalent maintenance loss cost based on the real-time load rate and efficiency values ​​of the auxiliary machine. The more severe the operating condition of the auxiliary machine, the higher the calculated maintenance cost, guiding the optimized scheduling to prioritize locking the auxiliary machine's operating condition within the high-efficiency operating zone.

[0049] Step S320: The environmental protection objective sub-function is constructed through three dimensions: carbon emission control, air pollutant compliance, and hidden costs of solid waste disposal. All parameters are dynamically priced based on carbon trading market conditions and environmental control strategies.

[0050] In a preferred embodiment of the present invention, based on the carbon emission control dimension, an indirect carbon emission cost breakdown is established. The auxiliary equipment's consumption of plant power will directly increase the unit's coal consumption. The system has a built-in formula for the unit's coal consumption characteristics under varying operating conditions. The increase in coal consumption is calculated based on the increase in the total plant power consumption. Then, the increase in carbon dioxide emissions is calculated by combining the low calorific value of coal and the carbon emission factor of coal. The system also captures the spot transaction price of the national carbon trading market in real time and converts the carbon emissions into the capital cost of purchasing carbon allowances.

[0051] In a preferred embodiment of the present invention, a penalty system for exceeding air pollutant standards is established based on the dimensions of air pollutant compliance, and real-time access is provided to the unit's CEMS flue gas online monitoring data and the NO emission data issued by the ecological environment bureau. xHourly emission limits for SO2 and particulate matter are established, along with a tiered penalty coefficient system. When the pollutant emission concentration derived from the optimized plan approaches the control threshold, a low-penalty early warning coefficient is activated. If the calculated emission value exceeds the legal limit, a high-level tiered penalty cost is automatically activated, with the penalty amount far exceeding the conventional energy-saving benefits. During Level I and Level II emergency control periods for heavy pollution weather, the penalty coefficients for all pollutants are increased by 2 to 5 times, and the optimized plan is forced to prioritize controlling coal consumption and reducing pollutant generation.

[0052] In a preferred embodiment of the present invention, based on the hidden costs of solid waste disposal, the operating costs of supporting environmental protection facilities are calculated. The power consumption of auxiliary equipment for desulfurization, dust removal, and denitrification is directly related to the consumption of reagents. The unit prices of raw materials such as limestone and ammonia are automatically updated and stored monthly. The increase in reagent consumption is calculated based on the changes in the load of auxiliary equipment and the cost is converted accordingly, thus fully covering the economic expenditures of the entire environmental protection chain.

[0053] Step S330: The technical safety target sub-function uses a graded penalty coefficient to quantify risk loss, and sets multi-level penalty items from four dimensions: main unit operation protection, health of individual auxiliary equipment, power quality of plant power grid, and hydraulic stability of the entire plant heating network.

[0054] In a preferred embodiment of the present invention, a supply and demand gap penalty item is set according to the host operation guarantee dimension. As the core host of the thermal power plant, the boiler and steam turbine have minimum guarantee thresholds for feedwater flow, primary air volume and induced draft volume. If the auxiliary machine scheduling scheme causes the medium supply to be lower than the minimum rigid parameter of the host, an ultra-large penalty coefficient is matched according to the supply and demand gap ratio. The larger the gap, the higher the penalty cost. From the objective function level, unreasonable scheduling schemes that threaten the safety of the host are directly eliminated.

[0055] In a preferred embodiment of the present invention, a penalty item for over-limit operating conditions of a single auxiliary machine is set according to the health dimension of a single auxiliary machine. The safe operating boundary of the equipment can be defined by relying on the real-time monitored bearing temperature, vibration spectrum, and inlet and outlet pressure parameters. When the optimized scheduling command causes the auxiliary machine to fall into dangerous operating condition ranges such as surge zone, cavitation zone, and over-temperature and overload zone, the equipment failure penalty cost is added step by step according to the deviation from the safe range.

[0056] In a preferred embodiment of the present invention, a power grid stability penalty is set according to the power quality dimension of the plant power grid. The simultaneous start-up and shutdown of multiple high-power auxiliary machines and sudden load changes can easily cause voltage drops and frequency deviations on the plant bus. The system embeds a plant impedance model to predict the bus voltage fluctuation amplitude caused by the control of the auxiliary machine cluster. If the voltage and frequency deviate from the national standard allowable range, a grid disturbance penalty is superimposed. The larger the fluctuation amplitude, the higher the penalty, thus ensuring the stability of the power quality of the plant system.

[0057] In a preferred embodiment of the present invention, a hydraulic stability penalty item for the cold and hot pipe networks is set according to the hydraulic stability dimension of the entire plant's heating network. When the optimization scheme causes the steam and hot water pipe network medium flow rate to exceed the limit or the pipe network pressure difference to change suddenly, the pipe network water hammer risk penalty is included to avoid the pipeline leakage hazard caused by drastic pressure fluctuations.

[0058] Step S340: Assign corresponding weight coefficients to the economic objective sub-function, environmental objective sub-function, and technical safety objective sub-function respectively, and form a quantifiable multi-objective optimization function through weighted fusion.

[0059] In a preferred embodiment of the present invention, the total objective function value can be expressed by the following formula. :

[0060] in, This represents the economic objective sub-function. This represents the environmental protection objective sub-function. This represents the sub-function representing the technical safety objective. This represents the weight coefficients of the economic objective sub-function. This represents the weight coefficients of the environmental protection objective sub-function. The weight coefficients of the technical safety objective sub-function are represented by the constraints. .

[0061] In this embodiment of the invention, the weighting coefficient is adaptively and dynamically adjusted through multi-dimensional triggering signals of the auxiliary equipment group of thermal power plants, and multi-level triggering preset thresholds and weight adjustment gradients are set to divide multiple control scenarios. The control scenarios may include peak electricity price control periods, heavy pollution weather graded early warning scenarios, peak summer and winter heating supply guarantee cycles, carbon price increases, planned maintenance windows of units, and monthly environmental emission assessment periods.

[0062] In a preferred embodiment of the present invention, when the peak electricity price of the power grid increases by more than 30% during the peak electricity price control period, the economic weight is applied. The tiered pricing structure is adjusted upwards, potentially reaching a maximum of 0.65, while the weightings for environmental protection and safety are proportionally reduced, shifting the optimization focus towards reducing plant power consumption and lowering electricity purchase costs. During off-peak hours with lower electricity prices, the economic weighting is reversed, prioritizing both environmental protection and equipment lifespan. For scenarios involving tiered early warnings of heavy pollution weather, a yellow environmental warning is issued. The environmental protection weight will be raised to 0.35 for orange alerts and 0.4-0.5 for red extreme alerts, with priority given to reducing coal consumption and pollutant emissions. For peak summer and winter heating supply periods, where there is a large power grid shortage and generating units bear the responsibility of providing backup heating, the technical safety weight will be adjusted accordingly. The weighting of environmental protection items has been increased to prioritize stable unit load and reliable equipment operation, eliminating safety hazards caused by extreme energy consumption reduction. In response to the sharp rise in carbon prices, with the carbon market carbon allowance price increasing by over 20% month-on-month, the weighting of environmental protection items has been automatically increased. Coal consumption and carbon allowance procurement expenditures have been reduced by optimizing auxiliary equipment operating conditions. During planned maintenance windows for units, the capacity of in-service units has decreased, and the margin of the power grid for plant use has declined. The weighting of safety has been increased to strictly control the disturbances caused by concentrated start-ups and shutdowns of high-power auxiliary equipment. For the monthly environmental ultra-low emission assessment period, with the monthly environmental data closing-loop assessment deadline approaching, the environmental weighting has been slightly increased to ensure compliance with emission standards for the entire month.

[0063] For example, a 660MW cogeneration unit operates during the daytime in summer, simultaneously under the dual control conditions of peak grid electricity price and orange heavy pollution warning from the ecological and environmental protection bureau. The system automatically completes the weight allocation according to preset trigger rules: economic weight 50%, environmental weight 40%, and equipment safety weight 10%. The system then breaks down the cost of each of the two operating schemes for the feedwater pump: Scheme A: Feedwater pump rated high-efficiency full-load operation: Electricity costs on the electricity consumption side are relatively high, increasing the economic component; however, the pump unit operates within the manufacturer's high-efficiency design range, with optimal overall hydraulic and motor efficiency, small increase in boiler coal consumption, low carbon emissions and pollutant output, and low environmental costs; all equipment operating parameters are within the safe range, with no safety penalty costs. Scheme B: Feedwater pump low-frequency reduced-capacity low-load operation: Auxiliary power consumption decreases significantly, electricity purchase costs are significantly reduced, and economic indicators are superior; however, long-term low-flow operation deviates from the high-efficiency range, causing a sharp decline in pump efficiency, increasing boiler coal consumption under the same feedwater parameters, and increasing CO2 and NO... x Simultaneous increases in environmental costs, coupled with the high environmental penalty coefficient under the orange alert, led to a significant surge in environmental costs. The operating conditions approached the cavitation critical line, triggering minor equipment safety penalties. After weighted overall objective function calculations, the total cost of scheme A was found to be far lower than that of scheme B, and the system automatically locked scheme A as the optimal operating strategy for this scheduling cycle.

[0064] Step S400: For the high nonlinearity and optimization characteristics of the multi-objective optimization function of the energy routing model, the objective decision variables are encoded with differentiated qubits through the principle of quantum superposition, and a hierarchical population initialization and adaptive quantum gate operator are designed to establish an observation decoding and hierarchical fitness evaluation process.

[0065] Please refer to Figure 5 In this embodiment of the invention, the target decision variable is encoded with differentiated qubits using the principle of quantum superposition, and a hierarchical population initialization and adaptive quantum gate operator are designed, which may include: Step S410: Using the classification and block-based qubit encoding method, the target decision variables in the energy routing model are mapped to the qubit probability amplitude space, and multiple encoding blocks are divided.

[0066] In a preferred embodiment of the present invention, multiple coding blocks may include three main coding blocks: a start / stop variable coding block, a load continuous variable coding block, and an energy routing selection coding block. Each quantum individual is composed of three coding bits concatenated together, and the bits in different blocks evolve independently without interference. For the discrete Boolean variable coding of auxiliary machine start / stop, the binary state of a single auxiliary machine's operation / stop is independently coded using a single quantum bit. Each auxiliary machine requiring start / stop optimization occupies one quantum bit. The start / stop variables of a batch of auxiliary machines are arranged consecutively to form start / stop coding blocks, facilitating subsequent unified quantum gate operations.

[0067] In a preferred embodiment of the present invention, for the segmented discrete encoding of the continuous load rate variable of auxiliary equipment, the continuous range of 0~100% rated load of auxiliary equipment is finely discretely divided into gears. The number of encoding bits is configured according to the equipment adjustment accuracy requirements. For high-precision variable frequency regulation auxiliary equipment, 5 qubits are selected, divided into 32 load gears. For baffle / hydraulic coupling coarse regulation auxiliary equipment, 3~4 qubits are selected, corresponding to 8 or 16 load gear ranges. Multiple qubits are combined to form binary codes. Different binary values ​​are mapped one-to-one to a fixed load percentage. The more bits, the higher the load adjustment resolution. The number of bits can be modified online according to the unit operation accuracy requirements. For the multi-energy flow routing selection variable encoding, for multi-path decision-making scenarios such as dual power supply of energy storage / external grid, direct steam supply / desuperheating and depressurization heating, and multi-source cooling water supply, corresponding qubits are configured according to the number of selectable paths. The bit combination state corresponds to different energy supply routes, realizing multi-path optimal encoding.

[0068] Step S420: Adopt a hierarchical differential superposition state initialization strategy to differentiate the initial values ​​of multiple coding blocks, so as to take into account both the common sense of the target thermal power plant equipment operation and the breadth of the whole domain search.

[0069] In a preferred embodiment of the present invention, the differentiated initial values ​​of multiple coding blocks may include differentiated initial values ​​of start-stop blocks, load blocks, and routing blocks. When initializing the start-stop coding block, based on the unit's daily operation log, the initial probability of the corresponding qubits of frequently operating key auxiliary equipment (e.g., main feedwater pumps and induced draft fans) is slightly increased to avoid blindly shutting down large areas during the initialization phase. The probability of standby auxiliary equipment remains superimposed, taking into account the possibility of standby operation search. When initializing the load coding block, the load level corresponding to the rated efficiency range of the auxiliary equipment is used as the reference center. The initial qubits are slightly tilted in probability near the reference level, which ensures that the algorithm prioritizes exploring efficient operating conditions while retaining equal exploration capability across the entire load range. When initializing the routing coding block, all routing bits are initialized to a completely equal probability superposition state, and the initial search probability of all energy transmission paths is consistent, with no initial path preference, achieving full coverage search of multiple routes.

[0070] Step S430: Using the quantum rotation gate as the global master evolution operator and the quantum NOT gate as the auxiliary mutation operator, an adaptive parameter adjustment rule is designed in conjunction with the convergence law of the target thermal power optimization, so that the operator parameters change dynamically with the iteration number and population diversity index.

[0071] In a preferred embodiment of the present invention, for the adaptive quantum rotation gate operator, the rotation gate is the core operator guiding the population towards the optimal solution. The magnitude and direction of the rotation angle are calculated in conjunction with the difference between the current individual's fitness and the global optimal individual's fitness. If the current individual's overall target cost is better than the population average, the qubit rotates in the positive direction towards the quantum state corresponding to the excellent solution, increasing the sampling probability of a high-quality scheduling scheme; if the current individual is worse than the average, the rotation probability amplitude is reversed, reducing the probability of a poor solution appearing.

[0072] In a preferred embodiment of the present invention, for the variable-probability quantum NOT gate mutation operator, the quantum NOT gate realizes random flipping of the qubit state to maintain population diversity and overcome the premature convergence dilemma of the algorithm. When the mutation probability is set to be linked to the iteration period, the mutation probability is maintained at a low value in the early stage of the algorithm iteration, during the global solution space exploration stage, to avoid frequent destruction of superior genes; in the middle and later stages of the iteration, when the population tends to be homogeneous, the mutation trigger probability is automatically increased, and some qubits are randomly flipped to escape the narrow interval of local optima.

[0073] In this embodiment of the invention, the establishment of the observation decoding and hierarchical fitness evaluation process may include quantum state observation collapse (e.g., following the quantum measurement principle to observe each quantum chromosome in the population bit by bit, randomly determining the value of 0 or 1 after collapse based on the probability amplitude of each bit, and generating a fixed binary string after all bits have been observed), partitioned reverse decoding (e.g., splitting the binary string according to the three major coding partitions of start / stop, load, and routing, and restoring it to the start / stop status of each auxiliary machine, the actual load percentage, and the energy supply path number, forming a complete and concrete scheduling scheme), preliminary screening of simplified constraints (e.g., after decoding, quickly verifying the auxiliary machine start / stop logic and the hard constraints of the upper and lower limits of the load, and directly assigning a high basic penalty to schemes with obvious violations, skipping the refined model calculation, and saving computing power) and refined fitness assignment (e.g., fully substituting the compliant scheduling scheme into the dynamic multi-objective total optimization function, calculating the comprehensive total cost of the scheme, and the cost value is the fitness index of the corresponding quantum individual, the lower the cost value, the better the individual performance).

[0074] For example, two key auxiliary equipment in the plant, an electric feedwater pump and a primary blower, were selected for algorithm implementation and debugging. Four qubits were allocated according to the coding rules to construct individual quantum entities. The first two bits were the feedwater pump load code, and the last two bits were the blower load code, with no additional routing variables. During population initialization, the feedwater pump load bits were slightly biased towards the high-efficiency load level, while the blower bits were superimposed with equal probability across all load levels, generating 50 initial quantum entities to form the initial population. During single observation and reverse decoding, after random quantum observation, the quantum state collapses to generate the binary code "1011". The first two "10" bits are decoded to obtain the feedwater pump load code. With the water pump at 50% rated load, the last two digits "11" are decoded to obtain the blower at 75% rated load, forming a concrete scheduling scheme. During the fitness quantification calculation, this load scheme is imported into the energy routing model and multi-objective function to calculate the economic, environmental, and safety costs, and sum them to obtain the comprehensive fitness score. When the quantum gate parameters are updated, it is found that the comprehensive cost of this scheme is better than the average fitness of the population. The quantum rotating gate adjusts the probability amplitude of the corresponding bit to increase the probability of generating the same high-quality load ratio again in subsequent iterations. When the homogeneity of the population is high in this round, the mutation probability of the quantum NOT gate is slightly increased to expand the search space for new working conditions.

[0075] Step S500: Based on the adaptive quantum gate operator, combined with the energy routing model and multi-objective optimization function, perform parallel global optimization in the ultra-high-dimensional hybrid solution space, cyclically execute the preset process, and generate the optimal scheduling scheme and multiple alternative schemes.

[0076] Please refer to Figure 6 In this embodiment of the invention, parallel global optimization is performed in the ultra-high-dimensional hybrid solution space, and a preset process is executed cyclically to generate an optimal scheduling scheme and multiple alternative schemes, which may include: Step S510: The adaptive quantum gate operator performs a coarse exploration of the entire domain in the early stage and a fine mining of local areas in the later stage. In the early stage, it traverses the ultra-high-dimensional mixed solution space to screen high-quality regions, and in the later stage, it locks the high-quality regions and finely adjusts the parameters.

[0077] Step S520: Generate the first generation of quantum populations in batches according to the rules of hierarchical population initialization. The number of the first generation of quantum populations is adaptively adjusted according to the dimension of the optimization variable.

[0078] In a preferred embodiment of the present invention, the initial quantum population can be generated in batches by calling the initialization rules. The number of individuals in the population is adaptively adjusted according to the dimension of the optimization variable. The population of small units (e.g., 300MW and below) is 40 / 50, and the population of large units of megawatt-level is expanded to 70 / 90. The entire initial population is stored in the parallel computing task queue and multi-threaded batch processing is enabled.

[0079] Step S530: Using a multi-threaded concurrent mode, perform quantum measurement collapse and binary decoding on all quantum individuals in the initial quantum population, and split them to obtain a scheduling scheme.

[0080] In a preferred embodiment of the present invention, a simplified hard constraint is used to quickly screen out extreme non-compliant schemes, while compliant schemes are connected in parallel to the energy routing model and multi-objective optimization function. The comprehensive fitness of each individual is calculated in a distributed and synchronous manner, and the cost data of all individuals are summarized to form a population fitness ranking table.

[0081] Step S540: Sort the target population according to the fitness value in a preset order, select the best quantum individual in the target population, and compare it with the historical best solution.

[0082] In a preferred embodiment of the present invention, the current generation of the population can be sorted in ascending order of fitness value to select the best individual of the current generation; the current best solution is compared with the historical global best solution item by item. If the new generation solution has a lower overall cost and complies with the constraints, it is used to replace and update the global best archive; if it is inferior to the historical best, the original global best data is retained. The top 5 second-best alternative solutions are stored synchronously in the alternative solution library.

[0083] Step S550: Based on the fitness of the optimal quantum individual and the overall population dispersion data, calculate the target parameters and generate offspring populations in batches.

[0084] In a preferred embodiment of the invention, the rotation angle of the quantum rotation gate and the mutation probability parameters of the quantum NOT gate can be automatically calculated based on the global optimal individual fitness and the overall population dispersion data. The updated quantum gate operator is then applied uniformly to all parent qubits to complete the iterative update of the quantum state, generating a new generation of quantum populations in batches, which then enter the next iteration queue. During iteration, population diversity indicators are monitored in real time, and the mutation probability is automatically increased when the individual fitness of the population is highly concentrated.

[0085] Step S560: Continuously execute the loop iteration. Stop the operation when any preset termination condition is met, and obtain the optimal scheduling scheme and multiple alternative schemes.

[0086] In a preferred embodiment of the present invention, the termination condition may include: the actual number of iterations reaching the system's preset maximum number of generations; the change in the global optimal objective function for N consecutive generations (e.g., 10 generations by default) being less than a preset minimum convergence threshold, and the parameters of the optimal solution essentially no longer changing, thus determining that the algorithm has converged. After convergence, the globally optimal scheduling scheme and the alternative scheme library are locked.

[0087] In this embodiment of the invention, to address various constraint violations in the decoded scheduling scheme, such as equipment output exceeding limits, ramp rate exceeding limits, pipeline transport capacity exceeding limits, and plant-wide power and heat supply-demand imbalance, a hierarchical external penalty function processing strategy can be adopted. This strategy categorizes constraints into two types based on their importance: hard constraints and soft constraints. Level 1 hard constraints (e.g., equipment rated upper and lower limits, minimum flow rate to prevent cavitation, pipeline rated capacity) involve equipment safety and system physical limits. Violations in these constraints incur high fixed penalty coefficients, resulting in penalties far exceeding the benefits of conventional optimization. This significantly deteriorates the fitness of violating individuals, leading to their rapid elimination by the algorithm during iteration. Level 2 soft constraints (e.g., minor ramp rate exceeding limits for auxiliary equipment, short-term minor supply-demand deviations) do not directly eliminate minor deviations from operating limits. Instead, a linearly increasing penalty proportional to the deviation magnitude is applied, with higher deviations resulting in higher additional costs. This guides the algorithm to gradually converge towards the compliant range, preserving a margin for error in minor adjustments. All penalty costs are uniformly added to the original multi-objective function result, achieving integrated calculation of constraints and optimization objectives.

[0088] In this embodiment of the invention, to prevent the loss of excellent scheduling schemes due to random mutations of quantum NOT gates during the iteration process, a three-level stratified elite retention system can be set up to ensure the stable inheritance of high-quality solutions: Level 1 elites: globally optimal individuals, the globally optimal quantum individuals are forcibly fully copied in each generation of iteration, do not participate in quantum gate mutation operations, and are unconditionally retained directly into the next generation of the population, thus preventing the loss of historical optimal solutions from the root; Level 2 elites: current top-performing individuals, the top 3% to 5% of the current population fitness are selected, their mutation trigger probability is slightly reduced, and their excellent genes are preferentially retained into the offspring; Level 3 elites: key schemes in the candidate pool, the top second-best solutions stored in the candidate scheme pool are injected into the population every fixed iteration period to enrich the pool of high-quality genes in the population and avoid the algorithm having a single convergence path.

[0089] For example, in the scenario of joint optimization of feedwater pumps and blowers for a 660MW unit, a population size of 50 and a maximum evolutionary iteration of 100 generations were set. A hierarchical penalty function and a three-level elite retention strategy were implemented to fully reproduce the iterative cycle changes: First generation (generation 1): The population was initially randomly distributed. Many individuals violated the hard constraint that the feedwater pump flow rate was lower than the minimum flow rate for cavitation prevention. The high penalty function in the first level increased the fitness cost, resulting in a high overall cost for the optimal solution, and the alternative solution library was empty. Iteration to generation 20: The quantum rotating door continuously guided the population towards a low-energy-consumption, high-efficiency range. The algorithm gradually avoided the low-flow-rate restricted areas of the pump units through penalty function feedback. The system gradually identified the pattern of a sharp drop in feedwater pump efficiency under long-term low load and an increase in overall cost instead of a decrease. It automatically raised the minimum operating load benchmark for the pump units, resulting in a precipitous decrease in the cost of the optimal solution. Iteration to generation 50: The algorithm autonomously explored a coordinated and efficient operation combination of "continuous fine-tuning of the feedwater pump frequency conversion + linkage of the blower blade opening". This combination scheme was selected as the global optimal archive, and three suboptimal schemes were simultaneously stored in the candidate library. Population diversity decreased slightly, and the system automatically increased the quantum mutation probability. Iteration to generation 90-100 convergence stage: After 10 consecutive generations, the fluctuation of the global optimal objective function was below the convergence threshold, the algorithm determined that convergence had terminated, output the optimal auxiliary machine load allocation instruction, and exported three backup scheduling schemes from the candidate library for flexible switching under abnormal operating conditions.

[0090] Step S600: Parse the optimal scheduling scheme and multiple alternative schemes into standardized control instructions, perform multiple security checks, and distribute and execute the standardized control instructions in a hierarchical manner, and provide execution status feedback and hierarchical handling of anomalies.

[0091] Please refer to Figure 7 In this embodiment of the invention, parsing the optimal scheduling scheme and multiple alternative schemes into standardized control instructions and performing multiple security checks may include: Step S610: Based on the target thermal power plant auxiliary equipment group database and the auxiliary equipment control hardware type, decompose the optimal scheduling scheme into instructions and generate control instructions in a differentiated manner according to the type of equipment.

[0092] In a preferred embodiment of the present invention, the differentiated generation of control commands for various equipment can include four types: frequency conversion type, baffle adjustment type, pneumatic adjustment type, and electrical switch type. For frequency conversion speed-regulated water pumps and condensate pumps, the optimal load percentage is combined with the frequency converter's V / F control curve to convert it into a target output frequency corresponding to a 4~20mA analog quantity, and the upper and lower limits of the frequency are synchronously limited to not exceed the rated operating range of the frequency converter. For blowers and induced draft fans that rely on hydraulic blades and inlet baffles to adjust the air volume, the valve characteristic curve is retrieved, and the load index is linearly converted into a baffle opening percentage command. For energy routing switching involving energy storage grid connection and plant power segmented switching, the circuit breaker, disconnector, and grounding switch opening and closing logic is split, and a single-step remote control opening and closing command sequence is generated, strictly following the step-by-step operation logic of the electrical operation ticket and eliminating one-step cross-logic operation.

[0093] Step S620: Add various tags (e.g., device number, operation sequence tag, operation allowed time window) to all control commands, distinguish between immediately executed commands and timed delayed execution commands (e.g., nighttime off-peak energy storage charging commands can be preset to trigger a timer), and obtain standardized control commands.

[0094] Step S630: Send standardized control commands to the preset security verification engine to perform multiple security verifications. If any verification step fails, an alarm will be issued and an error log will be stored.

[0095] In a preferred embodiment of the present invention, multiple safety verifications may include: a first-level electrical five-prevention logic verification, which retrieves the real-time energized status of the plant's primary power topology and the grounding switch position signal, and verifies each instruction for violations such as opening and closing disconnect switches under load, closing grounding switches while energized, accidentally entering energized intervals, and power supply parallel non-synchronization, and predicts short-circuit risks based on topology connectivity algorithms, directly and permanently blocking high-risk operations; a second-level auxiliary machine process interlocking condition verification, which links with real-time collected equipment status data to check the necessary start-up conditions such as the lubricating oil system, sealing water system, cooling system, minimum flow valve, and unit protection trip signal, and blocks start-up instructions if any auxiliary machine lacks any permissible start-up condition, and blocks all remote control operation instructions for equipment in fault trip or maintenance status at all times; The third level of cross-equipment command conflict verification involves horizontally comparing multiple commands from the same equipment within a short period to identify contradictory commands such as simultaneous start / stop and load exceeding the upper limit. Vertically, it verifies the collaborative control logic of multiple auxiliary machines. For example, it prohibits the full opening of the blower when the induced draft fan is completely off to prevent the furnace negative pressure from exceeding the limit and triggering the unit's protection action.

[0096] In a preferred embodiment of the present invention, after the command is issued, full-cycle feedback monitoring can be activated. The system sets differentiated feedback timeout thresholds according to equipment type. If there is no position feedback for 15 seconds for switch-type equipment, or if the actual frequency of frequency converter deviates from the target by more than 5% after 30 seconds, it is determined to be an execution abnormality. A three-level fault handling mechanism is established: Level 1: Minor deviation, where the actual operating condition deviates slightly from the command value, with no risk of exceeding limits. The system automatically makes small adjustments to correct the control output and continuously tracks parameter changes; Level 2: Moderate fault, where the actuator is stuck or the valve is not in position. The system suspends subsequent control commands in the same group, maintains the current operating condition of the equipment, and sends a pop-up warning to the host computer and pushes information to the operation duty terminal; Level 3: Severe fault, where parameters exceed limits or the equipment is at risk of over-temperature or over-pressure. The system immediately issues an emergency interlock / shutdown protection command, triggering the DCS safety interlock action to prevent the fault from escalating.

[0097] Step S700: Based on the deployed global sensor network, collect the operating data of the target thermal power plant auxiliary equipment group after executing standardized control commands. Using digital twin technology, compare the deviation between the predicted values ​​and the operating data, automatically correct the energy routing model parameters and adaptive quantum gate operator parameters, and realize closed-loop self-optimization of the entire system.

[0098] In this embodiment of the invention, when comparing the deviation between predicted values ​​and operational data using digital twin technology, the entire data acquisition link can be reused to collect actual measured data on auxiliary machine power consumption, output, pipeline flow, energy storage charging and discharging, and heat load after the current round of dispatch instructions has been implemented. The same set of dispatch instructions is substituted into the digital twin model to retrieve the theoretical predicted power consumption and output values ​​of the model, and the relative error between the predicted values ​​and the actual on-site operational data is calculated item by item. Based on mechanism analysis and big data clustering algorithms, the types of deviations are distinguished. Systematic overall constant deviation = equipment aging, pipeline renovation causing the original parameters of the model to fail; random and irregular small deviation = instantaneous meteorological disturbances, temporary fluctuations in external grid electricity prices, and other accidental interferences. Random disturbances do not require model correction, while systematic deviations initiate the parameter calibration process.

[0099] In this embodiment of the invention, if pumps and fans experience aging problems such as impeller wear, increased sealing gaps, and increased bearing losses during long-term operation, the original factory efficiency curve deviates from the actual operating conditions. The system refits the new efficiency characteristic curve of the equipment using a polynomial based on multi-cycle measured operating data, automatically replaces the old performance parameters in the model, and updates the equipment loss coefficient synchronously. After the completion of infrastructure renovations such as adding energy storage equipment, upgrading the heating network, and adding plant transformers in the plant area, the sensing network collects the measurement point data of the new equipment, and the model automatically adds topology nodes and transmission edges, completing the microgrid topology matrix update without manual intervention.

[0100] In this embodiment of the invention, the core control parameters of the algorithm can be dynamically modified based on historical iteration convergence performance to adapt to changes in field conditions: When the quantum rotation angle is adaptively adjusted, if the convergence speed of the algorithm is too slow and the optimization magnitude of the optimal solution is weak over multiple generations, the quantum rotation gate deflection step size is automatically increased to accelerate the convergence speed of population optimization. If the optimal solution frequently oscillates and fluctuates in the later stages of iteration, the rotation angle is reduced to improve convergence stability. When the mutation probability is dynamically controlled, if the individual solutions of the population are highly homogeneous over multiple generations and lose the ability to explore the solution space, the mutation trigger probability of the quantum NOT gate is automatically increased to improve population diversity and prevent the algorithm from getting stuck in a local optimum prematurely. After all model parameters and algorithm hyperparameters are calibrated, the system automatically saves the new version of the model file, waiting for the next scheduling cycle to trigger a new round of data acquisition, modeling optimization, and instruction issuance, forming a permanent self-optimization closed loop.

[0101] On the other hand, the present invention also provides a collaborative optimization operation system for auxiliary equipment groups of thermal power plants. The collaborative optimization operation system includes a control module, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor. The processor executes the computer program to implement the collaborative optimization operation method for auxiliary equipment groups of thermal power plants as described above.

[0102] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied 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.

[0103] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0104] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0105] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0106] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0107] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0108] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, 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), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0109] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0110] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A method for coordinated and optimized operation of auxiliary equipment groups in a thermal power plant, characterized in that, The collaborative optimization operation method includes: Based on the global sensor network deployed in the auxiliary equipment group of the target thermal power plant, multi-dimensional raw data is collected, and data preprocessing is performed through edge computing to generate a standardized dataset. Based on a standardized dataset and combined with the factory performance parameters of the auxiliary equipment group of the target thermal power plant, a multi-physics coupling model is established for each auxiliary equipment of the target thermal power plant, and a multi-layer constraint equation set is constructed to generate an energy routing model with complete constraints. Based on the energy routing model and standardized dataset, multiple cost items are separated and corresponding weight coefficients are configured. Through weighted fusion, a multi-objective optimization function that can be quantified and solved is formed. To address the high nonlinearity and multi-objective optimization characteristics of the energy routing model, the objective decision variables are encoded with differentiated qubits using the quantum superposition principle. A hierarchical population initialization and adaptive quantum gate operator are designed, and an observation decoding and hierarchical fitness evaluation process is established. Based on the adaptive quantum gate operator, combined with the energy routing model and multi-objective optimization function, parallel global optimization is performed in the ultra-high-dimensional hybrid solution space, and a preset process is executed cyclically to generate the optimal scheduling scheme and multiple sets of alternative schemes. The optimal scheduling scheme and multiple alternative schemes are parsed into standardized control instructions, which are then subjected to multiple safety checks. The standardized control instructions are then distributed and executed in a hierarchical manner, and the execution status is fed back and anomalies are handled in a hierarchical manner.

2. The method for coordinated and optimized operation of auxiliary equipment groups in thermal power plants according to claim 1, characterized in that, The global sensor network adopts a three-layer distributed three-dimensional deployment architecture, including an auxiliary equipment body perception layer, a plant microgrid topology perception layer, and an environment and external market perception layer. Based on the global sensor network deployed in the auxiliary equipment group of the target thermal power plant, it collects multi-dimensional raw data, including: Measurement points are deployed for the target energy-consuming auxiliary equipment of the target thermal power plant auxiliary equipment group, and the monitoring units are divided into three major monitoring dimensions: electrical, mechanical and fluid, as the sensing layer of the auxiliary equipment body; For the overall power grid and multi-energy flow pipeline topology of the target thermal power plant, data acquisition points are deployed on the power supply side, the heating and steam network side, and the energy storage side to construct the multi-energy production, transmission, and distribution links within the target thermal power plant area, serving as the plant area microgrid topology sensing layer and environment; External variable data are accessed from two dimensions: the natural environment and the policy market of the target thermal power plant area. The interference of environmental and policy changes on the optimization results is eliminated, which serves as the perception layer of environment and external market.

3. The method for coordinated and optimized operation of auxiliary equipment groups in thermal power plants according to claim 1, characterized in that, The process of establishing a multi-physics coupled model for each target thermal power plant auxiliary machine, constructing a multi-layered constraint equation set, and generating an energy routing model with complete constraints includes: Based on the theory of directed weighted graphs, the microgrid topology is decomposed into a set of multiple types of topological nodes and a set of multiple types of directed edges for energy transmission. Each node and directed edge for transmission is accompanied by corresponding attribute parameters to define the energy routing network topology. Based on the energy routing network topology, the full performance curves are digitally integrated, the drift mapping of variable operating conditions is performed, and two-dimensional state variables are defined to establish a multi-physics coupling model corresponding to each target thermal power plant auxiliary machine. Based on the multiphysics coupling model, a multi-layer constraint equation set is built from four dimensions: energy conservation, equipment limits, pipeline transmission limits, and plant supply and demand balance. This generates an energy routing model with complete constraints, ensuring that all scheduling schemes derived from the model meet the on-site physical operation criteria.

4. The method for coordinated and optimized operation of auxiliary equipment groups in thermal power plants according to claim 1, characterized in that, The multiple cost components include economic objective sub-functions, environmental objective sub-functions, and technical safety objective sub-functions. Based on the energy routing model and standardized dataset, multiple cost components are separated and assigned corresponding weight coefficients. These are then weighted and fused to form a quantifiable multi-objective optimization function, including: The economic objective sub-function is the sum of the direct production cost and the implicit equipment loss cost throughout the entire cycle, and is divided into three independent calculation modules: time-of-use plant power cost, auxiliary machine start-up and shutdown loss cost, and variable operating condition related maintenance cost. The environmental protection objective sub-function is constructed through three dimensions: carbon emission control, air pollutant compliance, and hidden costs of solid waste disposal. All parameters are dynamically priced based on carbon trading market conditions and environmental control strategies. The technical safety target sub-function uses a graded penalty coefficient to quantify risk loss, and sets up multi-level penalty items from four dimensions: main unit operation protection, health of individual auxiliary equipment, power quality of plant power grid, and hydraulic stability of plant heating network. Each of the economic objective sub-function, environmental objective sub-function, and technical safety objective sub-function is assigned a corresponding weight coefficient, and a multi-objective optimization function that can be quantified and solved is formed by weighted fusion.

5. The method for coordinated and optimized operation of auxiliary equipment groups in thermal power plants according to claim 4, characterized in that, The weighting coefficients are adaptively and dynamically adjusted through multi-dimensional triggering signals of the auxiliary equipment group of thermal power plants, and multi-level trigger preset thresholds and weight adjustment gradients are set to divide multiple control scenarios. The control scenarios include peak electricity price control periods, heavy pollution weather graded early warning scenarios, peak summer and winter heating supply guarantee cycles, carbon price increases, planned maintenance windows of units, and monthly environmental emission assessment periods.

6. The method for coordinated and optimized operation of auxiliary equipment groups in thermal power plants according to claim 1, characterized in that, The method of encoding the target decision variable with differentiated qubits using the principle of quantum superposition and designing hierarchical population initialization and adaptive quantum gate operators includes: A classification and block-based qubit encoding method is adopted to map the target decision variables in the energy routing model to the qubit probability amplitude space and divide it into multiple encoding blocks; A hierarchical differentiated superposition initialization strategy is adopted to distinguish the differentiated initial values ​​of multiple coding blocks, so as to take into account both the common sense of the operation of the target thermal power plant equipment and the breadth of the whole domain search. Using quantum rotation gate as the global master evolution operator and quantum NOT gate as the auxiliary mutation operator, an adaptive parameter adjustment rule is designed in conjunction with the convergence law of target thermal power optimization, so that the operator parameters change dynamically with the iteration number and population diversity index.

7. The method for coordinated and optimized operation of auxiliary equipment groups in thermal power plants according to claim 1, characterized in that, The process of parallel global optimization in the ultra-high-dimensional hybrid solution space, cyclically executing a preset process, and generating an optimal scheduling scheme and multiple alternative schemes includes: The adaptive quantum gate operator performs a coarse exploration of the entire domain in the early stage and a fine mining of local areas in the later stage. In the early stage, it traverses a large area of ​​the ultra-high-dimensional mixed solution space to select high-quality regions, and in the later stage, it locks the high-quality regions and finely adjusts the parameters. The first-generation quantum population is generated in batches according to the rules of hierarchical population initialization, and the number of the first-generation quantum population is adaptively adjusted according to the dimension of the optimization variable. A multi-threaded concurrent mode is adopted to perform quantum measurement collapse and binary decoding on all quantum individuals in the first generation of quantum population, and the scheduling scheme is obtained by splitting them; The target population is sorted according to its fitness value in a preset order, the best quantum individual in the target population is selected, and compared with the historical best solution. Based on the fitness of the optimal quantum individual and the overall population dispersion data, the target parameters are calculated, and offspring populations are generated in batches. The algorithm continuously executes a loop iteration, stopping the operation when any preset termination condition is met, thus obtaining the optimal scheduling scheme and multiple alternative schemes.

8. The method for coordinated and optimized operation of auxiliary equipment groups in thermal power plants according to claim 1, characterized in that, The process of parsing the optimal scheduling scheme and multiple alternative schemes into standardized control instructions and performing multiple security checks includes: Based on the database of auxiliary equipment groups of the target thermal power plant and the types of auxiliary equipment control hardware, the optimal scheduling scheme is decomposed into instructions, and control instructions are generated differently for each type of equipment. Add various tags to all control commands to distinguish between immediately executed commands and timed delayed execution commands, thus obtaining standardized control commands; Standardized control commands are sent to a preset security verification engine for multiple security checks. If any check fails, an alarm is issued and an error log is stored.

9. The method for coordinated and optimized operation of auxiliary equipment groups in thermal power plants according to claim 1, characterized in that, The collaborative optimization operation method also includes: Based on the deployed global sensor network, the system collects operational data of the target thermal power plant auxiliary equipment group after executing standardized control commands. Using digital twin technology, the system compares the deviation between the predicted values ​​and the operational data, automatically corrects the energy routing model parameters and adaptive quantum gate operator parameters, and achieves closed-loop self-optimization of the entire system.

10. A collaborative optimization operation system for auxiliary equipment groups in a thermal power plant, characterized in that, The collaborative optimization operation system includes a control module, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor. The processor executes the computer program to implement the collaborative optimization operation method for auxiliary equipment groups in thermal power plants according to any one of claims 1-9.

Images