Multi-objective energy saving and carbon reduction collaborative optimization method and system for commercial buildings

Abstract

The application discloses a multi-objective energy-saving and carbon-reducing collaborative optimization method and system for commercial buildings. The collaborative optimization method comprises the following steps: obtaining real-time building energy consumption data, environmental parameters, passenger flow density, electricity price information, power grid carbon intensity factor, equipment operating state and fire alarm signals; a multi-objective optimization model is constructed, and a multi-objective particle swarm optimization algorithm is used to obtain a Pareto optimal solution set; the multi-objective optimization model takes minimizing energy consumption cost, minimizing carbon emissions, minimizing customer experience deviation and minimizing operation safety risk as the target, takes building energy consumption system operating parameters as the decision variable, and takes carbon emission quota constraint, equipment operation safety constraint, customer experience bottom line constraint, fire safety mandatory constraint and guiding constraint as the constraint condition; and according to a safety, experience and energy-saving three-level progressive decision mechanism, an optimal control strategy is selected to be converted into equipment instructions and executed. The application can realize three-dimensional balance of energy-saving and carbon-reducing, customer experience and operation safety.

Description

Technical Field

[0001] This invention relates to the field of building carbon reduction optimization technology, and in particular to a multi-objective energy-saving and carbon reduction collaborative optimization method and system for commercial buildings. Background Technology

[0002] Commercial buildings (such as shopping malls, office buildings, and commercial complexes) are major consumers of urban energy, accounting for approximately 30% to 40% of total urban energy consumption. Lighting, air conditioning, and ventilation systems are the primary energy sources, accounting for over 70%. Commercial buildings need to achieve a comprehensive goal of energy conservation and carbon reduction, ensuring customer experience, and guaranteeing operational safety. However, there are significant conflicts among these three objectives: excessive energy conservation may lead to uncomfortable indoor temperature and humidity and insufficient lighting, affecting customer experience; a one-sided pursuit of energy conservation may ignore equipment operating limits, causing safety hazards; while excessively ensuring safety redundancy will increase energy consumption and carbon emissions.

[0003] Traditional building automation systems (BAS) are mostly based on fixed schedules or simple rules, lacking the intelligent optimization capability for dynamic balancing of multiple objectives. Existing optimization methods often simplify multi-objective problems into single-objective problems, making it difficult to handle conflicts and dynamics between objectives, and unable to adapt to uncertainties such as fluctuating customer flow, weather changes, and electricity price fluctuations in commercial buildings. More importantly, existing methods generally lack embedded modeling of operational safety constraints, and the optimization results may violate safety standards for fire protection, equipment, and other aspects. Summary of the Invention

[0004] The purpose of this invention is to address the shortcomings of traditional methods in terms of multi-objective dynamic balance, embedding of operational safety constraints, and adaptability to uncertainties. It provides a multi-objective energy-saving and carbon-reduction collaborative optimization method and system for commercial buildings that can efficiently handle the three-dimensional balance of energy saving and carbon reduction, customer experience, and operational safety. By incorporating the three dimensions of energy saving and carbon reduction, customer experience, and operational safety into a unified multi-objective optimization framework, it solves the technical problem of balancing these three aspects in the operation and management of commercial buildings.

[0005] According to one aspect of this application, a multi-objective energy-saving and carbon-reduction collaborative optimization method for commercial buildings is provided, comprising the following steps:

[0006] Acquire real-time building energy consumption data, environmental parameters, passenger flow density, electricity price information, grid carbon intensity factor, equipment operating status, and fire alarm signals; environmental parameters include temperature, humidity, CO2, and illuminance;

[0007] A multi-objective optimization model is constructed and solved using a multi-objective particle swarm optimization algorithm to obtain the Pareto optimal solution set. The multi-objective optimization model aims to minimize energy consumption cost, minimize carbon emissions, minimize customer experience deviation, and maximize operational safety margin. The operating parameters of the building energy consumption system are used as decision variables, and the constraints are carbon emission quota constraints, equipment operation safety constraints, customer experience baseline constraints, fire safety mandatory constraints, and guiding constraints.

[0008] Based on a three-tiered progressive decision-making mechanism of safety, experience, and energy conservation, the optimal control strategy is selected from the Pareto optimal solution set, converted into equipment instructions, and sent to the building control system for execution.

[0009] Preferably, the building energy consumption data includes at least the energy consumption data of the building's air conditioning system, lighting system, elevator system, fresh air unit system, and cold storage system, and the energy consumption data includes at least one of active power, reactive power, current, and voltage.

[0010] Preferably, the objective function of the multi-objective optimization model is expressed as follows:

[0011] ;

[0012] in This represents the energy cost target, indicating the energy consumption cost per unit of time. This indicates the carbon emission target, and also indicates the amount of carbon emissions per unit of time. This indicates the deviation from the customer experience target. Indicates operational safety risk objectives. Indicate the decision-making objective. The decision variables are formed by the target controllable parameters, including the target controllable parameters of each system and the delay time of each system. The target controllable parameters of each system include the dimming ratio of the lighting system, the set temperature and humidity of the air conditioning system, the number of elevators in operation and the start-stop frequency of the elevator system, the charging power of the cold storage system, and the operating frequency of the fresh air unit system.

[0013] Preferably, the energy cost target is determined by the time-of-use electricity price and the power of the decision variable, as follows:

[0014] ;

[0015] in, Representing decision variables At any moment power, Indicates the number of decision variables. Indicates time-of-use electricity pricing; Represents decision variables, Indicates time interval, Indicates the scheduling period;

[0016] The carbon emission target is determined by multiplying the energy consumption per unit time by the grid carbon intensity factor.

[0017] Preferably, the customer experience deviation target is determined by a weighted sum of the absolute values ​​of the differences between temperature, humidity, illuminance, and CO2 concentration and their respective set values, as shown below:

[0018] ;

[0019] in For temperature, For humidity, For illuminance, CO2 concentration , , , This indicates the set values ​​for temperature, humidity, illuminance, and CO2 concentration. To allow for fluctuation dead zones, , , , These are the weighting coefficients corresponding to the results calculated based on temperature, humidity, illuminance, and CO2 concentration; the smaller the deviation, the better the customer experience.

[0020] Preferably, the operational safety risk target is determined by the current value of the safety indicator and the safety threshold;

[0021] ;

[0022] in, For the first Current values ​​of each security indicator Indicates the number of indicators. As a safety threshold, As the normalization factor, Represents an exponential function; The smaller the value, the greater the operational safety margin, and the further the system is from the danger boundary.

[0023] Preferably, the constraints are divided into first-level constraints, second-level constraints, and third-level constraints;

[0024] The first-level constraints are hard constraints and cannot be violated. They include mandatory fire safety constraints and equipment operation safety constraints. The mandatory fire safety constraints include at least that smoke detectors, sprinklers, emergency lighting, and fire elevators should not be shut down for energy saving; in the event of a fire, the system should be forced to switch to fire mode, and the multi-objective energy saving and carbon reduction collaborative optimization instructions should be automatically invalidated. The equipment operation safety constraints include at least that the start-stop interval of the air conditioning compressor should not be less than a preset time threshold to avoid damage from frequent start-stops; when elevators are out of service during off-peak hours, at least one fire elevator should be kept running 24 hours a day and should have an automatic wake-up function.

[0025] The secondary constraints are soft constraints, including carbon emission quota constraints and customer experience baseline constraints; the carbon emission quota constraints include ≤ Carbon quota per unit time × (1 + ), To allow for over-emission coefficient, ≤5%; the customer experience baseline constraint includes The value is adjusted based on passenger flow density, with a maximum of 1.0.

[0026] The three levels of constraints are guiding constraints, including energy efficiency guiding constraints to encourage equipment to operate in the high-efficiency range, and green electricity priority constraints to prioritize the consumption of green electricity when it is available.

[0027] Preferably, the step of using a multi-objective particle swarm optimization algorithm to solve the multi-objective optimization model includes:

[0028] An adaptive inertial weight, determined by the upper and lower limits of particle swarm inertia weight, balances global exploration and local development capabilities. Dynamically adjusted volume learning factors and social dynamic learning factors are introduced to enhance the algorithm's adaptability in dynamic environments. Solutions that violate the first-level hard constraints are eliminated and not stored in the Pareto archive. When the size of the Pareto archive exceeds the upper limit, solutions with higher safety margins are retained based on crowding distance. Solutions that violate the second-level constraints are processed using a safety penalty function method, and finally, the Pareto optimal solution set is obtained.

[0029] Preferably, the selection of the optimal control strategy from the Pareto optimal solution set based on the three-level progressive decision-making mechanism of safety, experience, and energy saving, and the selection of the optimal control strategy by sequentially executing the three-level progressive decision-making mechanism of safety filtering, experience matching, and energy saving selection, includes:

[0030] The system employs a safety filtering decision to remove all solutions that may trigger safety risks from the Pareto optimal solution set; an experience matching decision to match corresponding customer experience preference weights based on the current time period and customer flow density; and an energy-saving optimization decision to select the solution with the lowest energy consumption cost among the remaining solutions that meet both safety and experience requirements, thus forming the optimal control strategy.

[0031] According to another aspect of this application, a multi-objective energy-saving and carbon-reduction collaborative optimization system for commercial buildings is provided, comprising:

[0032] The data acquisition module is used to acquire real-time energy consumption data, environmental parameters, passenger flow density, electricity price information, power grid carbon intensity factor, equipment operating status, and fire alarm signals; environmental parameters include at least temperature, humidity, CO2, and illuminance;

[0033] The model building and solution module is used to construct a multi-objective optimization model and solve the multi-objective optimization model using a multi-objective particle swarm optimization algorithm to obtain a Pareto optimal solution set. The multi-objective optimization model aims to minimize energy consumption cost, minimize carbon emissions, minimize customer experience deviation, and minimize operational safety risks. It uses the operating parameters of the building energy consumption system as decision variables and carbon emission quota constraints, equipment operation safety constraints, customer experience baseline constraints, fire safety mandatory constraints, and guiding constraints as constraints.

[0034] The strategy output module is used to select the optimal control strategy from the Pareto optimal solution set based on the three-level progressive decision-making mechanism of safety, experience, and energy saving, convert it into equipment instructions, and send them to the building control system for execution.

[0035] This invention incorporates energy conservation and carbon reduction, customer experience, and operational safety into a unified multi-objective optimization framework, achieving a three-dimensional objective balance and solving the pain point of difficulty in balancing these three aspects in the operation and management of commercial buildings. In the empirical application of the target experimental commercial complex, it achieved 17.3% energy saving in the air conditioning system while maintaining customer comfort satisfaction at over 90%, and no safety incidents caused by the energy-saving strategy occurred.

[0036] This invention implements an embedded design of safety constraints, establishes a three-level safety constraint system, and places hard constraints such as fire safety and equipment safety at the highest priority of the optimization algorithm. The optimization strategy is based on ensuring safety. It ensures that no optimization control strategy is at the expense of safety. At the same time, it establishes a safety monitoring module independent of the optimization algorithm, which can forcibly take over control and switch to fire mode in abnormal situations, ensuring that no optimization strategy exceeds the safety bottom line. The forced switching mechanism of safety mode provides the last line of defense.

[0037] This invention utilizes adaptive inertia weights and dynamic learning factors to efficiently solve multi-objective particle swarm optimization (EMOPSO) problems. Adaptive inertia weights are introduced to balance global exploration and local development capabilities, while dynamic learning factors enhance the algorithm's adaptability to dynamic environments such as passenger flow fluctuations and weather changes. A safety-priority Pareto archive set is constructed, enabling efficient solutions to large-scale multi-objective optimization problems. This results in fast convergence and high-quality solution sets for EMOPSO, allowing collaborative optimization decisions for thousands of devices to be completed within 15 seconds, meeting the real-time control requirements of commercial buildings.

[0038] This invention proposes a three-level progressive decision-making mechanism for safety, experience, and energy saving. It combines the real-time operating status of commercial buildings (such as time period type, passenger flow density, and real-time electricity price) to select the optimal control strategy from the Pareto optimal solution set step by step. The optimal control strategy is then sent to systems such as air conditioning, lighting, and elevators through an edge gateway for execution and to receive execution feedback for adjustment and optimization, forming an intelligent closed loop of perception, optimization, execution, and feedback.

[0039] This invention can dynamically adapt to uncertainties such as passenger flow fluctuations, weather changes, and electricity price fluctuations, and automatically adjust and optimize preferences under different operating conditions. It has strong adaptability and can achieve intelligent collaborative control at all times and in all scenarios. Attached Figure Description

[0040] Figure 1 This is a flowchart of the multi-objective energy-saving and carbon-reduction collaborative optimization method for commercial buildings according to the present invention.

[0041] Figure 2 This is a schematic diagram of the multi-objective energy-saving and carbon-reduction collaborative optimization system for commercial buildings according to the present invention. Detailed Implementation

[0042] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0043] See Figure 1 As shown in the exemplary embodiment of this application, the multi-objective energy-saving and carbon-reduction collaborative optimization method for commercial buildings is implemented by performing the following steps:

[0044] S1. Acquire real-time building energy consumption data, environmental parameters, passenger flow density, electricity price information, grid carbon intensity factor, equipment operating status, and fire alarm signals; environmental parameters include temperature, humidity, CO2, and illuminance;

[0045] S2. Construct a multi-objective optimization model and solve it using a multi-objective particle swarm optimization algorithm to obtain the Pareto optimal solution set. The multi-objective optimization model aims to minimize energy consumption cost, minimize carbon emissions, minimize customer experience deviation, and minimize operational safety risks. It uses the operating parameters of the building energy consumption system as decision variables and carbon emission quota constraints, equipment operation safety constraints, customer experience baseline constraints, fire safety mandatory constraints, and guiding constraints as constraints.

[0046] S3. Based on the three-level progressive decision-making mechanism of safety, experience, and energy saving, the optimal control strategy is selected from the Pareto optimal solution set, converted into equipment instructions, and sent to the building control system for execution.

[0047] As an example, in step S1, the data acquisition and processing can be performed using the following steps:

[0048] Real-time collection of multi-source operational data from commercial buildings (such as shopping malls and office complexes), including:

[0049] Energy consumption data: Active power, reactive power, current, and voltage of systems such as air conditioning, lighting, elevators, and ventilation are collected in 15-second increments using 0.5-second smart meters.

[0050] Environmental parameters: Temperature and humidity sensors (accuracy ±0.3℃, ±3%RH), CO2 concentration sensors (accuracy ±30ppm), and illuminance sensors (accuracy ±5%) were deployed in areas such as the atrium, shops, office areas, and parking lots to collect environmental parameter data.

[0051] Passenger flow density: Real-time passenger flow (people / hour) is obtained through entrance and floor heat map cameras, and aggregated by area into passenger flow density levels (e.g., high, medium, and low based on a certain threshold).

[0052] External dynamic data: real-time time-of-use electricity prices (peak price, valley price, flat price) are accessed from the power trading center, and dynamic carbon intensity factors (kg CO2 / kWh, refreshed every 15 minutes) are obtained from the grid dispatch.

[0053] Safety status data: signals from fire alarm systems (smoke detectors, heat detectors, manual alarms), equipment operating status (compressor start / stop count, motor temperature, bearing vibration), etc.

[0054] All collected data is timestamped via an edge gateway. Missing values ​​are filled using linear interpolation or forward padding. Outliers (exceeding 3) are handled accordingly. (Range) data is removed and marked. After preprocessing, the data is stored in a time-series database at 15-second granularities.

[0055] In step S2, the multi-objective optimization model can be established through the following steps:

[0056] A mathematical model was established using a typical 24-hour operating day for commercial buildings as the optimization cycle.

[0057] First, select multiple target controllable parameters covering the building's energy consumption system as corresponding decision variables, and form a decision variable vector X:

[0058] ;

[0059] in, Indicates the dimming ratio of the lighting system (0.5~1.0, corresponding to an illuminance of 50~750lx, emergency lighting ≥50lx, automatic restart upon power failure); This indicates the set temperature (24-26℃ in summer, 18-22℃ in winter, fluctuation ≤ ±2℃, to avoid discomfort) and humidity (40%-65% RH, to avoid condensation or excessive dryness) of the air conditioning system. This indicates the operating frequency of the fresh air handling unit (20-50Hz, corresponding to a CO2 concentration of 600-1000ppm in the air, for forced ventilation in densely populated areas). This indicates the number of elevators in operation (1 to 8 units, dynamically adjusted, with at least one fire elevator operating 24 hours a day, accounting for 0.2 to 1.0% of the total). This indicates the cooling capacity of the cold storage system (0-100% of the design capacity). This indicates the dimming ratio of parking lot lighting (0.3 to 1.0). This indicates the delay time for each system (2-30 minutes, varying by system).

[0060] Secondly, a multi-objective function is constructed to achieve an optimal balance between energy conservation and carbon reduction, customer experience, and operational safety, and a minimized objective vector is constructed:

[0061] ;

[0062] Energy cost target (Yuan / day):

[0063] ;

[0064] In the formula, Decision variables At any moment Power, time interval =15min=0.25h, It uses time-of-use pricing, with a day divided into 24 hours and 15-minute intervals, totaling 96 intervals.

[0065] carbon emission targets (kgCO2 / day): ,in Represents the dynamic grid carbon intensity factor (real-time access). This indicates energy consumption, which is the instantaneous power consumption per unit time.

[0066] Customer experience deviation (Dimensionless):

[0067] ;

[0068] In the formula, , , , These are the measured temperature, humidity, illuminance, and CO2 concentration, respectively. =1℃ is the allowable fluctuation dead zone, and the weighting coefficient is dynamically adjusted according to passenger flow density (the experience weight is increased by 20% when there is high passenger flow).

[0069] operational safety margin (Dimensionless, aiming to maximize is equivalent to minimizing the negative safety margin):

[0070] ;

[0071] in For the first Current values ​​of safety indicators (such as compressor exhaust temperature, motor vibration amplitude). As a safety threshold, This is the normalization factor. The smaller the value, the greater the operational safety margin.

[0072] Set constraints (divided into three levels according to priority):

[0073] Among them, the first-level constraints (hard constraints, which must not be violated) include:

[0074] Firefighting equipment (smoke detectors, sprinklers, emergency lighting, and fire elevators) must not be shut down or downgraded due to energy conservation directives; the start-stop interval of air conditioning compressors must be ≥5 minutes; at least one fire elevator must be kept running 24 hours a day during off-peak hours; the minimum cold storage capacity reserve of the cold storage system must be ≥80% of the design load (during continuous high-temperature weather warnings).

[0075] Secondary constraints (soft constraints, which can be violated to a limited extent, with penalties added for violations) include:

[0076] ≤ Carbon quota per unit time × (1 + ), To allow for excess emission coefficients, the unit of time can be calculated on a daily basis. For example, carbon emissions ≤ daily carbon allowance × (1 + 5%); customer experience deviation target. ≤0.5 (business hours) and not less than 0, or 1.0 (non-business hours).

[0077] Level 3 constraints (guiding constraints, encouraging their fulfillment) include:

[0078] Equipment should be operated in its most efficient range (e.g., 40%–80% load rate for chiller units); green electricity should be used preferentially when available.

[0079] When using the particle swarm optimization algorithm to solve the problem in this application, the following steps are performed:

[0080] Algorithm initialization:

[0081] Set particle swarm size For example, 50 represents the maximum number of iterations. , such as 200; the position vector of the particle composed of decision variables. The velocity vector is randomly generated within the feasible region of the decision variables. Initialize to zero; initialize Pareto archive set. Empty.

[0082] Introducing adaptive inertia weights:

[0083] Make inertial weight With the number of iterations Linearly decreasing, balancing global search and local exploitation: current iteration number The inertia weight is calculated as follows:

[0084] ;

[0085] in This represents the current iteration number. Indicates inertia weight, , These represent the maximum and minimum weight values, respectively. =0.9, =0.4.

[0086] Introducing a dynamic learning factor strategy for individual learning factors Social learning factors Dynamic adjustment:

[0087] For example, such as ;

[0088] Initial individual learning factors The scale is relatively large, so exploration is encouraged; learning factors will be implemented later. Larger size accelerates convergence. This indicates the maximum number of iterations.

[0089] During the solution process, the particle velocity and position update formulas are as follows:

[0090] ;

[0091] in It is a random number. This represents the particle's own historical optimal position. The global leader is randomly selected from the Pareto archive in the current iteration. Indicates the first The particles at the current iteration number The speed of the drop Indicates the first The particles at the current iteration number The lower position; This indicates the updated particle velocity. Represents the individual learning factor, taking , Represents the social learning factor, taking .

[0092] The constraint handling and Pareto archive maintenance during particle algorithm solution are as follows:

[0093] Level 1 Constraint Filtering: New Particles If any hard constraint is violated, the device is discarded and reinitialized.

[0094] Second-level constraint penalty: For particles that violate second-level constraints, a penalty term is added to the original objective function value.

[0095] ,

[0096] in This is a penalty coefficient, such as 1000, which is dynamically adjusted according to the security level. To constrain the amount of violations, This represents the objective function for increasing the penalty.

[0097] Pareto Dominance Judgment:

[0098] Compare the penalized target vectors; add non-dominated solutions to the Pareto archive set. At the same time, the dominated solution is removed.

[0099] Archive set size control: If the number of Pareto archive sets... We use crowding distance sorting to retain solutions with more even distribution and higher safety margins. This indicates the Pareto archive set size threshold, such as 30.

[0100] When the rate of change of the archive set is less than 1% for several consecutive generations (e.g., 20 generations), or when the maximum number of iterations is reached, the algorithm terminates and outputs the Pareto optimal solution set.

[0101] In this application, after outputting the Pareto optimal solution set, a three-level progressive decision-making strategy is implemented to screen for the optimal control strategy. Generally, based on the real-time operating status of the commercial building, the final control strategy is selected from the Pareto optimal solution set. The three-level progressive decision-making includes:

[0102] Safety filtering decision: Eliminate any solutions in the Pareto optimal solution set that may lead to safety warnings (such as equipment temperature exceeding limits, false alarms from fire protection systems, etc.).

[0103] Experience matching decision: Dynamically adjust the experience weighting coefficient based on the current customer flow density (high, medium, low) and time of day (open, closed): e.g., high customer flow + open time: experience weighting For example, if there is low customer traffic and the business is not operating: the experience weight is reduced by 50%, and energy saving is given priority.

[0104] Energy-saving optimization decision-making: Selecting the energy cost target from the solution set that meets safety and user experience requirements. The lowest / minimum solution is used as the final control strategy.

[0105] The final control strategy is converted into BACnet, MODBUS and other protocol commands through the edge gateway and sent to field controllers such as air conditioning, lighting, and elevators for execution.

[0106] The above three-level progressive decision-making is carried out sequentially. After the safety filtering decision, the experience matching decision is made from the remaining candidate solution set that meets the safety decision. The relevant experience weights are adjusted according to the current passenger flow density, etc. The solution set that meets the customer experience deviation is selected from the remaining candidate solution set that meets the safety decision. Then, the solution with the minimum energy consumption cost target is selected from the solution set that meets the customer experience deviation as the final control strategy.

[0107] In some implementations, after obtaining the Pareto optimal solution set, the Pareto set is first trimmed using a dual-mode approach based on equipment status, safety envelope, and fire safety confidence score to obtain a safe solution set; then, experience weights are calculated and determined based on passenger flow density and time period, and candidate solutions that meet the customer experience deviation criteria are selected from the safe solution set; finally, the solution with the minimum comprehensive cost is selected from the candidate solution set based on a weighted average of real-time electricity price and grid carbon intensity. Specifically, future... If the control sequence corresponding to the Pareto optimal solution causes the predicted temperature rise to cross the safety envelope (equipment dynamic temperature threshold), then the Pareto optimal solution is discarded.

[0108] When performing equipment condition simulation and defining the safety envelope, the equipment safety temperature is denoted as... The formula is used to determine the result: ;

[0109] Indicates the rated temperature on the equipment nameplate. Indicates the aging coefficient. This indicates the ambient temperature correction factor; the higher the ambient temperature, the narrower the safety margin.

[0110] ;

[0111] in, Indicates the cumulative number of start-stop operations. Indicates the number of start-stop cycles during the design life. This represents a coefficient, ranging from 0.15 to 0.25.

[0112] ;

[0113] in, Represents the coefficient. Indicates the current ambient temperature. This indicates the design ambient temperature.

[0114] The temperature is predicted using a thermal model as follows:

[0115] ;

[0116] in, Indicates the predicted temperature. Indicates the current temperature. Indicates ambient temperature. Indicates thermal resistance. This indicates the device load rate corresponding to the solution. This represents the cooling time constant.

[0117] Based on the above temperature predictions and the equipment temperature over-limit threshold, the following adjustments are made:

[0118] like, If so, then the corresponding solution is removed.

[0119] in, It is a safety margin buffer coefficient (taken as 0.05 to 0.1), which intercepts in advance rather than waiting for the line to be crossed before intercepting.

[0120] In one embodiment, when cropping based on fire safety confidence scores, the following steps can be used:

[0121] ;

[0122] in, This indicates the fire safety confidence score for the preset area. Indicates a smoke detector signal. Indicates the temperature signal, This indicates a manual alarm signal. Indicates the direction of passenger flow occupancy. , , , This represents the corresponding weighting coefficient, which sums to 1. It is a binary or probability signal (0 / 1 or 0~1 confidence level). If... The solutions involved are removed from the Pareto set, and the remaining solutions are used as solutions for subsequent customer experience decision-making processes.

[0123] In some embodiments, a customer experience threshold is set in the customer experience decision-making process. After security filtering based on the customer experience threshold, a solution set is selected, and a solution that meets the customer experience criteria is selected, thereby realizing the selection of the solution set in the customer experience decision-making process.

[0124] In some implementations, the method further includes a security mode forced switching mechanism:

[0125] When a fire alarm signal is triggered, the system unconditionally switches to fire mode or safety mode, emergency lighting is forcibly turned on, all equipment except fire elevators is shut down, non-fire power supplies are cut off, the air conditioning system is shut down (except for smoke exhaust fans), elevators are forced to land on the first floor and are deactivated, the optimization algorithm output is automatically disabled, an alarm is sent to the operation and maintenance center, and an event log is recorded. When the equipment operating parameters approach the safety threshold, the penalty coefficient of the corresponding constraint is automatically increased to ensure that the safety boundary is not exceeded.

[0126] This application also includes a safety monitoring mechanism independent of the multi-objective optimization algorithm, which monitors fire alarm signals and the operating status of critical equipment in real time, and performs safety monitoring and forced switching. Unconditionally, it switches to safety mode when any of the following conditions occur:

[0127] Any fire alarm (smoke detector, manual alarm) is triggered; critical equipment parameters (such as motor current, compressor exhaust temperature) exceed the warning threshold; or communication is interrupted for more than 30 seconds.

[0128] Furthermore, the technology of this application feeds back data such as actual energy consumption, carbon emissions, user experience satisfaction, and safety incidents to the cloud database after each control command is executed, achieving closed-loop feedback and model self-optimization. For example, it can be set to automatically retrain the prediction model on historical data weekly, using incremental learning to update the multi-GBDT load prediction model and particle swarm optimization parameters. This can be achieved by comparing the effects of the old and new strategies through A / B testing. If the overall performance improvement of the new strategy exceeds 5%, the new strategy will be automatically deployed to the production environment.

[0129] Example:

[0130] Take a 100,000㎡ commercial complex (including a shopping mall, office buildings, and parking lots) as an example.

[0131] On a typical summer workday, the outdoor temperature is 35°C and the humidity is 70%.

[0132] Passenger flow density: Medium from 10:00 to 12:00 on weekdays, and high from 14:00 to 17:00;

[0133] Electricity price: 1.2 yuan / kWh during peak hours (10:00-15:00, 18:00-21:00), 0.8 yuan / kWh during normal hours (7:00-10:00, 15:00-18:00, 21:00-23:00), and 0.4 yuan / kWh during off-peak hours (23:00-7:00).

[0134] Grid carbon intensity factor: 0.65 kg CO2 / kWh during peak hours, 0.55 during normal hours, and 0.45 during off-peak hours.

[0135] The algorithm converged after 150 iterations, yielding a Pareto front containing 18 non-dominated solutions.

[0136] Based on the principle of prioritizing user experience during peak hours, the solution is selected. .

[0137] After implementation, the air conditioning system achieved an energy saving rate of 17.3%, an overall energy saving rate of 15.2%, and reduced carbon emissions by 18.6 tons / day; customer comfort satisfaction reached 92.7%; and there were no safety violations.

[0138] Performance Comparison: Compared with traditional fixed schedule control, the method of the present invention has significant improvements in three dimensions: energy saving, carbon reduction, and user experience, as shown in Table 1.

[0139] Table 1

[0140] .

[0141] According to embodiments of this application, a multi-objective energy-saving and carbon-reduction collaborative optimization system for commercial buildings is also provided, see [link to relevant documentation]. Figure 2 As shown, it includes:

[0142] The data acquisition module is used to acquire real-time energy consumption data, environmental parameters, passenger flow density, electricity price information, power grid carbon intensity factor, equipment operating status, and fire alarm signals; environmental parameters include at least temperature, humidity, CO2, and illuminance. The model building and solution module is used to construct a multi-objective optimization model and solve it using a multi-objective particle swarm optimization algorithm to obtain a Pareto optimal solution set. The multi-objective optimization model aims to minimize energy consumption cost, minimize carbon emissions, minimize customer experience deviation, and minimize operational safety risks. It uses the operating parameters of the building energy consumption system as decision variables and carbon emission quota constraints, equipment operation safety constraints, customer experience baseline constraints, fire safety mandatory constraints, and guiding constraints as constraints. The strategy output module is used to select the optimal control strategy from the Pareto optimal solution set according to the three-level progressive decision-making mechanism of safety, experience, and energy saving, convert it into equipment instructions, and send them to the building control system for execution.

[0143] Furthermore, this application also includes a safety monitoring module independent of the optimization algorithm, used for real-time monitoring of fire alarm signals and the operating status of critical equipment, performing safety monitoring and forced switching. Unconditionally switching to safety mode occurs when one of the following conditions is met:

[0144] Any fire alarm (smoke detector, manual alarm) is triggered; critical equipment parameters (such as motor current, compressor exhaust temperature) exceed the warning threshold; or communication is interrupted for more than 30 seconds.

[0145] In safe mode, execute the predetermined emergency strategy: emergency lighting is forcibly turned on, non-fire-fighting power is cut off, the air conditioning system is shut down (except for the smoke exhaust fan), the elevator is forced to land on the first floor and is stopped, and an alarm is sent to the operation and maintenance center and the event log is recorded.

[0146] The system in this application uses technology to feed back data such as actual energy consumption, carbon emissions, user experience satisfaction, and safety incidents to a cloud database after each control command is executed, achieving closed-loop feedback and model self-optimization. For example, if set to automatically retrain the multi-objective optimization model weekly using historical data, and update the multi-objective optimization model and particle swarm optimization parameters using incremental learning, the effects of the new and old strategies can be compared through A / B testing. If the overall performance improvement of the new strategy exceeds 5%, the new strategy will be automatically deployed to the production environment.

[0147] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. It will be apparent to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and that the present invention can be implemented in other specific forms without departing from the spirit or basic characteristics of the present invention. Therefore, the embodiments should be regarded as exemplary and non-limiting in all respects. The scope of the present invention is defined by the appended claims rather than the foregoing description, and therefore all changes falling within the meaning and scope of the equivalents of the claims are intended to be included within the present invention.

[0148] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A multi-objective collaborative optimization method for energy conservation and carbon reduction in commercial buildings, characterized in that, include: Acquire real-time building energy consumption data, environmental parameters, passenger flow density, electricity price information, grid carbon intensity factor, equipment operating status, and fire alarm signals; environmental parameters include temperature, humidity, CO2, and illuminance; A multi-objective optimization model is constructed and solved using a multi-objective particle swarm optimization algorithm to obtain the Pareto optimal solution set. The multi-objective optimization model aims to minimize energy consumption cost, minimize carbon emissions, minimize customer experience deviation, and minimize operational safety risk. The operating parameters of the building energy consumption system are used as decision variables, and the constraints are carbon emission quota constraints, equipment operation safety constraints, customer experience baseline constraints, fire safety mandatory constraints, and guiding constraints. Based on a three-tiered progressive decision-making mechanism of safety, experience, and energy conservation, the optimal control strategy is selected from the Pareto optimal solution set, converted into equipment instructions, and sent to the building control system for execution.

2. The multi-objective energy-saving and carbon-reduction collaborative optimization method for commercial buildings according to claim 1, characterized in that, The building energy consumption data includes at least the energy consumption data of the building's air conditioning system, lighting system, elevator system, fresh air unit system, and cold storage system. The energy consumption data includes at least one of active power, reactive power, current, and voltage.

3. The multi-objective energy-saving and carbon-reduction collaborative optimization method for commercial buildings according to claim 1, characterized in that, The objective function of the multi-objective optimization model is expressed as follows: ； in This represents the energy cost target, indicating the energy consumption cost per unit of time. This indicates the carbon emission target, and also indicates the amount of carbon emissions per unit of time. This indicates the deviation from the customer experience target. Indicates operational safety risk objectives. Indicate the decision-making objective. The decision variables are formed by the target controllable parameters, including the target controllable parameters of each system and the delay time of each system. The target controllable parameters of each system include the dimming ratio of the lighting system, the set temperature and humidity of the air conditioning system, the number of elevators in operation and the start-stop frequency of the elevator system, the charging power of the cold storage system, and the operating frequency of the fresh air unit system.

4. The multi-objective energy-saving and carbon-reduction collaborative optimization method for commercial buildings according to claim 3, characterized in that, The energy cost target is determined by the time-of-use electricity price and the power of the decision variable, as follows: ； in, Representing decision variables At any moment power, Indicates the number of decision variables. Indicates time-of-use electricity pricing; Represents decision variables, Indicates time interval, Indicates the scheduling period; The carbon emission target is determined by multiplying the energy consumption per unit time by the grid carbon intensity factor.

5. The multi-objective energy-saving and carbon-reduction collaborative optimization method for commercial buildings according to claim 3, characterized in that, The customer experience deviation target is determined by a weighted sum of the absolute values ​​of the differences between temperature, humidity, illuminance, and CO2 concentration and their respective set values, as shown below: ； in, For temperature, For humidity, For illuminance, CO2 concentration , , , This indicates the set values ​​for temperature, humidity, illuminance, and CO2 concentration. To allow for fluctuation dead zones, , , , These are the weighting coefficients corresponding to the results calculated based on temperature, humidity, illuminance, and CO2 concentration; the smaller the deviation, the better the customer experience.

6. The multi-objective energy-saving and carbon-reduction collaborative optimization method for commercial buildings according to claim 3, characterized in that, The operational safety risk target is determined by the current value of the safety indicator and the safety threshold. ； in, For the first Current values ​​of each security indicator Indicates the number of indicators. As a safety threshold, As the normalization factor, Represents an exponential function; The smaller the value, the greater the operational safety margin, and the further the system is from the danger boundary.

7. The multi-objective energy-saving and carbon-reduction collaborative optimization method for commercial buildings according to claim 1, characterized in that, The constraints are divided into first-level constraints, second-level constraints, and third-level constraints; The first-level constraints are hard constraints and cannot be violated. They include mandatory fire safety constraints and equipment operation safety constraints. The mandatory fire safety constraints include at least that smoke detectors, sprinklers, emergency lighting, and fire elevators should not be shut down for energy saving; in the event of a fire, the system should be forced to switch to fire mode, and the multi-objective energy saving and carbon reduction collaborative optimization instructions should be automatically invalidated. The equipment operation safety constraints include at least that the start-stop interval of the air conditioning compressor should not be less than a preset time threshold to avoid damage from frequent start-stops; when elevators are out of service during off-peak hours, at least one fire elevator should be kept running 24 hours a day and should have an automatic wake-up function. The secondary constraints are soft constraints, including carbon emission quota constraints and customer experience baseline constraints; the carbon emission quota constraints include ≤ Carbon quota per unit time × (1 + ), To allow for over-emission coefficients; the customer experience baseline constraints include Adjustments will be made based on passenger flow density. The three levels of constraints are guiding constraints, including energy efficiency guiding constraints to encourage equipment to operate in the high-efficiency range, and green electricity priority constraints to prioritize the consumption of green electricity when it is available.

8. The multi-objective energy-saving and carbon-reduction collaborative optimization method for commercial buildings according to claim 7, characterized in that, The method of using a multi-objective particle swarm optimization algorithm to solve the multi-objective optimization model includes: An adaptive inertial weight, determined by the upper and lower limits of particle swarm inertia weight, balances global exploration and local development capabilities. Dynamically adjusted volume learning factors and social dynamic learning factors are introduced to enhance the algorithm's adaptability in dynamic environments. Solutions that violate the first-level constraints are eliminated and not stored in the Pareto archive. When the Pareto archive exceeds the upper limit, solutions with higher safety margins are retained based on crowding distance. Solutions that violate the second-level constraints are processed using a safety penalty function method, and finally, the Pareto optimal solution set is obtained.

9. The multi-objective energy-saving and carbon-reduction collaborative optimization method for commercial buildings according to claim 8, characterized in that, The aforementioned three-tiered progressive decision-making mechanism based on safety, user experience, and energy efficiency selects the optimal control strategy from the Pareto optimal solution set. This selection is achieved by sequentially executing the three-tiered progressive decision-making mechanism of safety filtering, user experience matching, and energy efficiency optimization, including: The system employs a safety filtering decision to remove all solutions that may trigger safety risks from the Pareto optimal solution set; an experience matching decision to match corresponding customer experience preference weights based on the current time period and customer flow density; and an energy-saving optimization decision to select the solution with the lowest energy consumption cost among the remaining solutions that meet both safety and experience requirements, thus forming the optimal control strategy.

10. A multi-objective energy-saving and carbon-reduction collaborative optimization system for commercial buildings, characterized in that: include: The data acquisition module is used to acquire real-time energy consumption data, environmental parameters, passenger flow density, electricity price information, power grid carbon intensity factor, equipment operating status, and fire alarm signals; environmental parameters include at least temperature, humidity, CO2, and illuminance; The model building and solution module is used to build a multi-objective optimization model and solve it using a multi-objective particle swarm optimization algorithm to obtain a Pareto optimal solution set. The multi-objective optimization model aims to minimize energy consumption cost, minimize carbon emissions, minimize customer experience deviation, and minimize operational safety risks. It uses the operating parameters of the building energy consumption system as decision variables and carbon emission quota constraints, equipment operation safety constraints, customer experience baseline constraints, fire safety mandatory constraints, and guiding constraints as constraints. The strategy output module is used to select the optimal control strategy from the Pareto optimal solution set based on the three-level progressive decision-making mechanism of safety, experience, and energy saving, convert it into equipment instructions, and send them to the building control system for execution.

Images