Building construction site energy-saving power utilization system and control method thereof

Abstract

The invention discloses a building construction site energy-saving power utilization system and a control method thereof, and belongs to the technical field of power utilization management. The system comprises an energy monitoring module, a load prediction module, a multi-objective optimization module, a dynamic scheduling module, a carbon efficiency evaluation module, an execution feedback module and a distributed power supply module. According to the method, data such as equipment vibration and rotating speed are acquired through the industrial bus to construct the equipment working condition vector, and the equipment working condition state vector and the temperature and humidity correction function are fused, so that the dynamic accurate prediction of the construction load is realized; the problem of large matching error caused by the fact that environmental parameters and process dynamic characteristics are not considered in building construction site load prediction in the prior art is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power management, in particular to an energy-saving power system for a construction site and a control method thereof. BACKGROUND

[0002] In the field of energy management of construction sites, load prediction is the core link to achieve energy-saving scheduling. Traditional technologies usually use historical data statistics or static time series models for load prediction. Such methods mainly construct prediction models based on historical energy consumption data of equipment operation, without considering environmental parameters and dynamic characteristics of construction processes in the modeling system. For example, in a high-temperature environment in summer, the cooling system energy consumption of construction machinery will increase significantly, and the traditional model often leads to a large deviation between predicted load and actual demand due to the lack of temperature and humidity correction mechanism.

[0003] Another limitation of the prior art is the lack of adaptability to dynamic changes in construction processes. Construction has typical stage characteristics, and the equipment configuration and operation mode of different processes differ significantly. However, traditional prediction models mostly use fixed parameters for modeling, which cannot dynamically adjust load prediction parameters with process switching, resulting in a sharp increase in prediction errors during process conversion, and further causing scheduling mismatch between distributed power and grid power purchase, leading to energy waste or reduced power supply reliability. SUMMARY

[0004] The embodiments of the present application provide an energy-saving power system for a construction site and a control method thereof, which solves the problem of large matching error in load prediction of a construction site in the prior art due to the lack of consideration of environmental parameters and dynamic characteristics of processes. The device working condition vector is constructed by collecting data such as device vibration and speed through an industrial bus, and the device working condition vector and temperature and humidity correction function are fused, thereby achieving dynamic and accurate prediction of construction load.

[0005] The embodiments of the present application provide an energy-saving power system for a construction site, which is used to implement an energy-saving power control method for a construction site, and includes an energy monitoring module, a load prediction module, a multi-objective optimization module, a dynamic scheduling module, a carbon efficiency evaluation module, an execution feedback module, and a distributed power module.

[0006] The energy monitoring module is used to collect photovoltaic power generation, grid power purchase, energy consumption of electrical equipment, and carbon emission intensity of the grid in real time.

[0007] The load prediction module is used to obtain a working condition vector, a construction equipment temperature, and an environmental humidity, and analyze total load demand of the site.

[0008] The multi-objective optimization module is configured to acquire real-time energy data, generate a distributed power output plan and a power grid electricity purchase strategy according to total site load demand and the real-time energy data, and take the minimum total electricity cost and the minimum carbon emission as targets;

[0009] The dynamic scheduling module is configured to decompose controllable device-level control instructions according to the distributed power output plan and the total site load demand;

[0010] The carbon efficiency evaluation module is configured to obtain a real-time carbon efficiency index according to the distributed power output plan and the power grid electricity purchase strategy;

[0011] The execution feedback module is configured to compare an instruction execution deviation and a carbon efficiency change amount, and trigger adaptive correction of an optimization model parameter;

[0012] The distributed power supply module is configured to execute control instructions of controllable devices at a current time and feed back actual output.

[0013] Further, the total site load demand acquisition step comprises:

[0014] The working condition state vector of the construction equipment j at the time t is constructed by collecting construction equipment vibration sensor and motor speed meter data through an industrial bus interface;

[0015] The process code, construction equipment temperature and environmental humidity are obtained from a BIM database, and a future period load prediction value is calculated through an energy consumption mapping function;

[0016] The energy consumption mapping function is:

[0017]

[0018] wherein, is the future period load prediction value of the construction equipment j, S j is the working condition state vector of the construction equipment j at the time t, Ψ j is a 1×n energy consumption mapping coefficient vector of the construction equipment j, k θ is a device temperature sensitivity coefficient, k W is an environmental humidity sensitivity coefficient, θ j is a device temperature, θ0 is a device temperature reference value, W t is an environmental humidity, W0 is an environmental humidity reference value;

[0019] Finally, the future period load prediction values of the devices are summarized as the total site load demand through an industrial Ethernet

[0020] Further, the working condition state vector of the construction equipment j at the time t is acquired in the following manner:

[0021] S j(t) = Φ · S j (t-1) + Γ · u j (t) + ω j ;

[0022] where S j (t) is the working condition state vector of construction equipment j at time t, S j (t-1) is the working condition state vector of construction equipment j at time t-1, u j (t) is the quantized control instruction of construction equipment j at time t, Φ is an n x n state transition matrix, Γ is an n x 1 control input matrix, the quantized control instruction u j (t) has an impact on the state, ω j is a Gaussian noise vector subject to

[0023] Further, the step of obtaining the distributed power output plan and the grid power purchase strategy includes:

[0024] A target function is established with the lowest total power consumption cost and the smallest carbon emission as the target:

[0025]

[0026] where F is the value of the target function, p elec (t) is the real-time electricity price, t is the time index, T is the total duration, λ(t) is the dynamic grid carbon intensity factor, P grid (t) is the grid power purchase amount, α is the economic cost weight, β is the carbon emission weight, and α + β = 1;

[0027] The constraint conditions of the target function are set:

[0028]

[0029] where m is the index number of the distributed power source, M is the total number of distributed power sources in the construction site, P dg,m (t) is the actual output power of the mth distributed power source at time t, is the total load demand of the site at time t, and ε(t) is the dynamic elastic margin coefficient;

[0030] Finally, the Pareto optimal solution set is solved by a particle swarm algorithm to generate the distributed power output plan and the grid power purchase strategy

[0031] Further, the dynamic elastic margin coefficient is calculated as:

[0032] The load change acceleration is obtained according to the total load demand of the site at time t; ​

[0033] Obtain the control instruction execution deviation in the past 24 hours, and obtain the variance thereof;

[0034] The dynamic elastic margin coefficient is obtained through a dynamic elastic margin coefficient formula

[0035]

[0036] Wherein, κ is a load change sensitivity coefficient, v is a control deviation sensitivity coefficient, δ u (τ) is the control instruction execution deviation at time τ, τ is an integral time variable, and the value range is [t-24h, t].

[0037] Further, the step of decomposing the controllable device-level control instruction comprises:

[0038] Obtain the partial derivative of the distributed power output plan with respect to the control instruction and the partial derivative of the site total load demand with respect to the control instruction, and construct a weighted Euclidean norm;

[0039] Wherein, W is a weight matrix, W = diag(w1,..., w K ) is a diagonal weight matrix, w k is the carbon efficiency sensitivity weight of controllable device k;

[0040] Obtain the smoothing term ρ|u k (t)-u k (t-1)|

[0041] The optimal control instruction solving function is obtained according to the weighted Euclidean norm and the smoothing term:

[0042]

[0043] Wherein, is the optimal control instruction of controllable device k at time t, is the distributed power output plan at time t, is the site total load demand at time t, u k (t) is the control instruction of controllable device k at time t, ρ is a control instruction smoothing coefficient, and u k (t-1) is the control instruction value at the previous time t-1.

[0044] Further, the carbon efficiency sensitivity weight of controllable device k is obtained according to the following formula:

[0045]

[0046] Wherein, λ(t) is a dynamic power grid carbon emission intensity factor, is the distributed power output plan of the mth distributed power at time t, A power grid electricity purchasing strategy.

[0047] Further, the real-time carbon efficiency index is obtained by the following formula:

[0048]

[0049] is the optimal carbon efficiency index at time t, reflecting the carbon emission efficiency of distributed power supply, is the distributed power output plan of the mth distributed power at time t, and M is the total number of distributed powers; λ(t) is a dynamic power grid carbon emission intensity factor at time t, is the power grid electricity purchasing strategy at time t.

[0050] Further, the dynamic power grid carbon emission intensity factor obtaining step comprises:

[0051] The real-time current of the power grid node is collected by a current transformer, the line resistance and the power grid topology distance are read from a nonvolatile memory, the dynamic power grid carbon emission intensity factor is calculated by 24-hour sliding window integration, and the calculation formula of the dynamic power grid carbon emission intensity factor is:

[0052]

[0053] wherein τ is an integral time variable, γ grid is the power grid marginal carbon emission intensity at time τ, P grid is the actual power grid electricity purchasing amount at time τ.

[0054] Further, the step of executing the control instruction of the controllable device at the current time comprises the following steps:

[0055] The line impedance is obtained by an impedance measuring device, and the real-time line loss rate ζ line (t) is calculated.

[0056] ζ line (t) = 1-cosφ V-I (t).

[0057] wherein φ V-I (t) is the voltage-current phase difference at time t;

[0058] Based on the upper and lower limits of the output are set;

[0059] wherein is the minimum output limit of the distributed power, is the optimal photovoltaic output plan at time t, is the optimal diesel generator output plan at time t, is the maximum output limit of the distributed power, ​is the load forecast value at time t;

[0060] Finally, the control instructions of the controllable device at the current moment are executed through the solid-state relay, and the trip protection is triggered when the upper limit is exceeded.

[0061] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages:

[0062] 1. The equipment vibration, speed and other data are collected through the industrial bus to construct the equipment operating condition vector, and the equipment operating condition state vector is integrated with the temperature and humidity correction function, thereby realizing dynamic and accurate prediction of construction load. This effectively solves the problem of large matching errors caused by the failure to consider environmental parameters and process dynamic characteristics in the existing construction site load prediction technology.

[0063] 2. By calculating the dynamic carbon emission intensity factor through 24-hour sliding window integration, the spatiotemporal changes of the grid's carbon emission intensity can be tracked in real time, thereby achieving dynamic quantification of the grid's carbon emissions from electricity purchases, and providing a real-time carbon efficiency benchmark for multi-objective optimization. This effectively solves the problem in existing technologies that carbon emission indicators cannot reflect the real-time carbon intensity of grid electricity.

[0064] 3. Through multi-objective Pareto optimization combined with particle swarm optimization, electricity costs and carbon emissions are simultaneously optimized and a solution set is generated, thereby achieving coordinated optimization of the "carbon-electricity" costs of the construction site, and then providing economic and low-carbon trade-off solutions for different construction stages. This effectively solves the problem in existing technologies that single-objective optimization cannot take into account both economic benefits and environmental protection requirements. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 This is a structural diagram of the energy-saving electricity system for a construction site provided in an embodiment of the present application. DETAILED DESCRIPTION

[0066] The embodiments of the present application solve the problem of large matching errors in load prediction of construction sites in the prior art due to failure to consider environmental parameters and dynamic characteristics of the process by providing an energy-saving electricity consumption system and control method for construction sites. The equipment operating condition vector is constructed by collecting data such as equipment vibration and speed through the industrial bus, and the equipment operating condition state vector and temperature and humidity correction function are integrated to achieve dynamic and accurate prediction of construction load.

[0067] In order to better understand the above technical solution, the above technical solution will be described in detail below with reference to the accompanying drawings and specific implementation methods.

[0068] like Figure 1As shown, the embodiment of the present application provides an energy-saving electricity system for a construction site, which is used to implement an energy-saving electricity control method for a construction site, including an energy monitoring module, a load forecasting module, a multi-objective optimization module, a dynamic scheduling module, a carbon efficiency evaluation module, an execution feedback module, and a distributed power supply module;

[0069] The energy monitoring module is used to collect the photovoltaic power generation P in the construction site in real time. pv (t), Power purchase quantity P grid (t), energy consumption of electrical equipment L i (t) and grid carbon emission intensity γ grid (t);

[0070] The load forecasting module is used to obtain the working state vector, construction equipment temperature and ambient humidity, and analyze the total load demand of the site.

[0071] The steps of obtaining the total site load demand include:

[0072] The data of the vibration sensor and motor tachometer of the construction equipment are collected through the industrial bus interface to construct the working state vector of the construction equipment j at time t;

[0073] S j (t)=Φ·S j (t-1)+Γ·u j (t)+ω j ;

[0074] Among them, S j (t) is the working state vector of construction equipment j at time t, S j (t-1) is the working state vector of construction equipment j at time t-1, u j (t) is the quantized control instruction of construction equipment j at time t, Φ is the n×n state transfer matrix, which describes the natural evolution law of equipment working conditions, Γ is the n×1 control input matrix, and the quantized control instruction u j (t) Impact on state, ω j For construction equipment j obey Gaussian noise vector distributed, Σ is an n×n symmetric positive definite matrix, whose diagonal elements Σ ii represents the variance of the i-th component in the operating state vector, and its off-diagonal elements Σ ij (i≠j) represents the covariance between the i-th component and the j-th component in the operating state vector, Used to describe the state vector S of the construction equipment working condition j Random interference in (t);

[0075] The process code, construction equipment temperature and environment humidity are obtained from the BIM database, and the future period load prediction value is calculated through the energy consumption mapping function;

[0076] The energy consumption mapping function is:

[0077]

[0078] wherein, is the future period load prediction value of the construction equipment j, Ψ j is the 1×n energy consumption mapping coefficient vector of the construction equipment j, k θ is the equipment temperature sensitive coefficient, k W is the environment humidity sensitive coefficient, θ j is the equipment temperature, θ0 is the equipment temperature reference value, W t is the environment humidity, W0 is the environment humidity reference value;

[0079] Finally, the future period load prediction values of each equipment are summarized as the site total load demand through industrial Ethernet

[0080] The multi-objective optimization module is used to obtain real-time energy data, and according to the site total load demand and the real-time energy data, a distributed power output plan and a power grid power purchase strategy are generated with the lowest total electricity cost and the smallest carbon emission as the target

[0081] The distributed power output plan and the power grid power purchase strategy are generated The obtaining step includes:

[0082] A target function with the lowest total electricity cost and the smallest carbon emission as the target is established:

[0083]

[0084] wherein, F is the target function value, p elec (t) is the real-time electricity price (yuan / kWh), t is the time index, T is the total duration, λ(t) is the dynamic power grid carbon intensity factor, P grid (t) is the power grid power purchase amount, λ(t)P grid (t) is the power grid power purchase carbon emission amount (kgCO2), α is the economic cost weight, β is the carbon emission weight, and α+β=1;

[0085] Secondly, the constraint conditions of the target function are set:

[0086]

[0087] where m is the index number of the distributed power supply, M is the total number of the distributed power supply in the construction site, P dg,m (t) is the actual output power of the mth distributed power supply at time t, is the total load demand of the site at time t, and ε(t) is a dynamic elastic margin coefficient.

[0088] Finally, the particle swarm algorithm is used to solve the Pareto optimal solution set to generate the distributed power supply output plan

[0089] and the grid power purchase strategy

[0090] When the particle swarm algorithm is used to solve the Pareto optimal solution set, the distributed power supply output plan and the grid power purchase strategy are encoded as particle positions, the particle swarm size and other parameters are set, and the positions and velocities are randomly initialized. The constraints are handled by a penalty function to ensure Then, the objective function value (total electricity cost and carbon emissions) of each particle is calculated, and the Pareto solution set is selected according to the non-dominated relationship and stored in the archive. Then, the particle velocity and position are updated (the velocity is guided by the individual optimal and global optimal), and after the update, the projection is performed on the feasible region, and the crowding distance is used to maintain the diversity of the solution set. Finally, when the iteration ends or the solution set is stable, the non-dominated solutions in the archive are output, and the and

[0091] The dynamic elastic margin coefficient is calculated as follows:

[0092] According to the total load demand of the site at time t, the load change acceleration is obtained

[0093] The control instruction execution deviation of the past 24 hours is obtained, and the variance Var[δ u (τ)] is obtained.

[0094] The dynamic elastic margin coefficient is obtained by the dynamic elastic margin coefficient formula

[0095]

[0096] where κ is the load change sensitivity coefficient, ν is the control deviation sensitivity coefficient, δ u (τ) is the control instruction execution deviation at time τ, and τ is the integral time variable, which is in the range of [t-24h, t].

[0097] The load change sensitivity coefficient K and the control deviation sensitivity coefficient v are obtained by training historical data in the construction phase: using historical construction data such as load change, control deviation and actual power supply margin demand in the foundation construction, main body construction and other stages, the least square method or machine learning algorithm is used to fit the optimal K and v values in different construction stages, and a phased coefficient library is formed for real-time calling.

[0098] The dynamic scheduling module is configured to decompose controllable device-level control instructions u k (t) according to the distributed power output plan and the total load demand of the site.

[0099] The step of decomposing the controllable device-level control instructions includes:

[0100] Obtaining the partial derivative of the distributed power output plan to the control instruction And the partial derivative of the total load demand of the site to the control instruction Respectively representing the influence degree (kW / control unit) of the control instruction on the power output and the load matching, and constructing a weighted Euclidean norm

[0101] Wherein, W is a weight matrix, W = diag(w1,..., w K ) is a diagonal weight matrix, w k is the carbon efficiency sensitivity weight of controllable device k, and diag(·) represents constructing a diagonal matrix with vector elements as diagonal elements.

[0102] Obtaining a smoothing term ρ|u k (t)-u k (t-1) to prevent the controllable device from frequently starting and stopping.

[0103] According to the weighted Euclidean norm and the smoothing term, an optimal control instruction solving function is obtained:

[0104]

[0105] Wherein, is the optimal control instruction of controllable device k at time t, is the distributed power output plan at time t, is the total load demand of the site at time t, u k (t) is the control instruction of controllable device k at time t, ρ is the control instruction smoothing coefficient, and u k (t-1) is the control instruction value at the previous time t-1.

[0106] The formula for obtaining the carbon efficiency sensitivity weight w k of the controllable device k is:

[0107]

[0108] wherein λ(t) is a dynamic grid carbon intensity factor, is the optimal output plan of the mth distributed power supply at time t, is the grid power purchase strategy.

[0109] represents the sensitivity of the control instruction to the carbon efficiency index, and the denominator is the square of the carbon efficiency index denominator item.

[0110] The carbon efficiency evaluation module is configured to obtain a real-time carbon efficiency index according to the optimal output plan and the grid power purchase strategy.

[0111]

[0112] wherein is the optimal carbon efficiency index at time t, reflecting the carbon emission efficiency of the distributed power supply, is the optimal output plan (kW) of the mth distributed power supply at time t, M is the total number of distributed power supplies; λ(t) is a dynamic grid carbon intensity factor (kgCO2 / kWh) at time t, is the grid power purchase strategy (kW) at time t;

[0113] The dynamic grid carbon intensity factor λ(t) acquisition step comprises:

[0114] The real-time current of the grid node is collected by a current transformer, the line resistance and the grid topology distance are read from a non-volatile memory, the dynamic grid carbon intensity factor is calculated by 24-hour sliding window integration, and the calculation formula of the dynamic grid carbon intensity factor is:

[0115]

[0116] wherein τ is an integral time variable (τ ∈ [t-24h, t]), γ grid (τ) is the grid marginal carbon intensity (kgCO2 / kWh) at time τ, grid (τ) is the actual grid power purchase amount (kW) at time τ.

[0117] wherein the numerator is the total carbon emission of the grid power purchase in the past 24 hours (kgCO2), the denominator is the total grid power purchase amount (kWh) in the same period, and the dynamic grid carbon intensity factor λ(t) reflects the rolling average carbon intensity of the grid power (kgCO2 / kWh);

[0118] The execution feedback module is configured to compare the instruction execution deviation δ u (t) with the carbon efficiency change amount Δη c (t), and trigger adaptive correction of the optimization model parameters;

[0119] The distributed power supply module is used to execute the control instructions u of the controllable device at the current moment k (t) and feedback actual output

[0120] The execution of the control instruction of the controllable device at the current moment comprises the following steps:

[0121] Obtain line impedance through impedance measurement device and calculate real-time line loss rate ζ line (t):

[0122] ζ line (t)=1-cosφ V-I (t)

[0123] Among them, φ V-I (t) is the voltage and current phase difference at time t (radians), ζ line (t) is the real-time line loss rate;

[0124] Then based on Set the upper and lower limits of output, where: is the minimum output limit of distributed power generation (kW), is the optimal photovoltaic output plan at time t (kW), is the optimal diesel generator output plan at time t (kW), is the maximum output limit of distributed power generation (kW), is the load forecast value at time t (kW);

[0125] in is the maximum available power (kW) taking into account line losses;

[0126] Finally, the control instructions of the controllable device at the current moment are executed through the solid-state relay, and the trip protection is triggered when the upper limit is exceeded.

[0127] In summary, the embodiment of the present application collects equipment vibration, speed and other data through the industrial bus to construct the equipment operating condition vector, and integrates the equipment operating condition state vector with the temperature and humidity correction function, thereby realizing dynamic and accurate prediction of construction load.

[0128] It will be understood by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0129] The present application is described in reference to the flowchart illustrations and / or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the 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, create means for implementing the functions specified in the flowchart illustrations and / or block diagrams block or blocks. Figure 1 one or more flowcharts and / or blocks Figure 1 one or more flowcharts and / or blocks

[0130] These computer program instructions can also be stored in a computer- readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instructions which implement the function specified in the flowchart illustrations and / or block diagrams block or blocks. Figure 1 one or more flowcharts and / or blocks Figure 1 one or more flowcharts and / or blocks

[0131] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart illustrations and / or block diagrams block or blocks. Figure 1 one or more flowcharts and / or blocks Figure 1 one or more flowcharts and / or blocks

[0132] While the preferred embodiments of the application have been described, additional variations and modifications can be made to the embodiments by those of skill in the art once they have the benefit of the present disclosure without departing from the spirit and scope of the application. Accordingly, the attached claims are intended to cover all such variations and modifications as falling within the scope of the application.

[0133] Obviously, numerous modifications and variations of the present application are possible in light of the above teachings. It is therefore to be understood that within the scope of the appended claims and their equivalents, the application can be practiced otherwise than as specifically described.

Claims

1. The energy-saving electricity system for construction sites is characterized by: Including energy monitoring module, load forecasting module, multi-objective optimization module, dynamic scheduling module, carbon efficiency evaluation module, execution feedback module, and distributed power supply module; The energy monitoring module is used to collect real-time data on photovoltaic power generation, power purchases from the power grid, energy consumption of electrical equipment, and carbon emission intensity of the power grid within the construction site; The load prediction module is used to obtain the working state vector, construction equipment temperature and ambient humidity, and analyze the total load demand of the site; The multi-objective optimization module is used to obtain real-time energy data and generate a distributed power output plan and grid power purchase strategy based on the total load demand of the site and the real-time energy data, with the goal of minimizing total electricity cost and carbon emissions; The dynamic scheduling module is used to decompose controllable device-level control instructions based on the distributed power output plan and the total load demand of the site; The carbon efficiency evaluation module is used to obtain real-time carbon efficiency indicators based on the distributed power output plan and the power grid purchase strategy; The execution feedback module is used to compare the instruction execution deviation with the carbon efficiency change, triggering the adaptive correction of the optimization model parameters; The distributed power supply module is used to execute the control instructions of the controllable devices at the current moment and feedback the actual output.

2. The control method for the energy-saving electricity system of a construction site according to claim 1, characterized in that: The steps of obtaining the total site load demand include: The data of the vibration sensor and motor tachometer of the construction equipment are collected through the industrial bus interface to construct the working state vector of the construction equipment j at time t; Obtain process codes, construction equipment temperature, and ambient humidity from the BIM database, and calculate load forecast values ​​for future periods using the energy consumption mapping function; The energy consumption mapping function is: in, is the load forecast value of construction equipment j in the future period, S j (t) is the working state vector of construction equipment j at time t, Ψ j is the 1×n energy consumption mapping coefficient vector of construction equipment j, kθ is the equipment temperature sensitivity coefficient, k W is the environmental humidity sensitivity coefficient, θ j is the device temperature, θ0 is the device temperature reference value, W t is the ambient humidity, W0 is the ambient humidity reference value; Finally, the load forecast values ​​of each device in the future period are summarized as the total load demand of the site through industrial Ethernet 3. The control method for the energy-saving electricity system of a construction site according to claim 2, characterized in that: The method of obtaining the working state vector of the construction equipment j at time t is as follows: S j (t)=Φ·S j (t-1)+Γ·u j (t)+ω j ; Among them, S j (t) is the working state vector of construction equipment j at time t, S j (t-1) is the working state vector of construction equipment j at time t-1, u j (t) is the quantized control instruction of construction equipment j at time t, Φ is the n×n state transfer matrix, Γ is the n×1 control input matrix, and the quantized control instruction u j (t) Impact on state, ω j For construction equipment j obey Distributed Gaussian noise vector.

4. The control method for the energy-saving electricity system of a construction site according to claim 1, characterized in that: The steps of generating a distributed power generation output plan and obtaining a power grid purchase strategy include: Establish an objective function with the goal of minimizing total electricity costs and carbon emissions: Among them, F is the objective function value, p elec (t) is the real-time electricity price, t is the time index, T is the total duration, λ(t) is the dynamic grid carbon emission intensity factor, P grid (t) is the amount of electricity purchased from the power grid, α is the economic cost weight, β is the carbon emission weight, and α+β=1; Set the constraints of the objective function: Where m is the index number of the distributed power source, M is the total number of distributed power sources in the construction site, P dg,m (t) is the actual output power of the mth distributed generation at time t, is the total site load demand at time t, ε(t) is the dynamic elastic margin coefficient; Finally, the particle swarm algorithm is used to solve the Pareto optimal solution set and generate the distributed power output plan. and power grid purchasing strategies 5. The control method for the energy-saving electricity system of a construction site according to claim 4, characterized in that: The dynamic elastic margin coefficient is calculated as: The load change acceleration is obtained based on the total load demand of the site at time t; Obtain the control instruction execution deviation in the past 24 hours and obtain its variance; The dynamic elastic margin coefficient is obtained by the dynamic elastic margin coefficient formula Among them, κ is the load change sensitivity coefficient, ν is the control deviation sensitivity coefficient, δ u (τ) is the control instruction execution deviation at time τ, τ is the integral time variable, and its value range is [t-24h, t].

6. The control method for the energy-saving electricity system of a construction site according to claim 1, characterized in that: The step of decomposing the controllable device-level control instructions includes: Obtain the partial derivatives of the distributed generation output plan with respect to the control command and the partial derivatives of the total site load demand with respect to the control command, and construct a weighted Euclidean norm; Where W is the weight matrix, W=diag(w1,...,w K ) is the diagonal weight matrix, w k is the carbon efficiency sensitivity weight of controllable device k; Get the smoothing term ρ|u k (t)-u k (t-1)|; The optimal control instruction solution function is obtained based on the weighted Euclidean norm and smoothness term: in, is the optimal control instruction of controllable device k at time t, is the distributed generation output plan at time t, is the total load demand of the site at time t, u k (t) is the control command of the controllable device k at time t, ρ is the control command smoothing coefficient, u k (t-1) is the control command value at the previous time t-1.

7. The control method for the energy-saving electricity system of a construction site according to claim 6, characterized in that: The formula for obtaining the carbon efficiency sensitivity weight of the controllable device k is: Among them, λ(t) is the dynamic grid carbon emission intensity factor, is the distributed power output plan of the mth distributed power source at time t, Power purchasing strategy for the grid.

8. The control method for the energy-saving electricity system of a construction site according to claim 1, characterized in that: The formula for obtaining the real-time carbon efficiency index is: in is the optimal carbon efficiency index at time t, reflecting the carbon emission efficiency of distributed power supply. is the distributed power output plan of the mth distributed power source at time t, M is the total number of distributed power sources; λ(t) is the dynamic grid carbon emission intensity factor at time t, is the power purchasing strategy of the power grid at time t.

9. The control method for the energy-saving electricity system of a construction site according to claim 1, characterized in that: The step of obtaining the dynamic grid carbon emission intensity factor includes: The real-time current of the grid nodes is collected through current transformers, the line resistance and grid topology distance are read from the non-volatile memory, and the dynamic grid carbon emission intensity factor is calculated by integrating the 24-hour sliding window. The calculation formula of the dynamic grid carbon emission intensity factor is: Among them, τ is the integration time variable, γ grid (τ) is the marginal carbon emission intensity of the power grid at time τ, P grid (τ) is the actual amount of electricity purchased from the power grid at time τ.

10. The control method for the energy-saving electricity system of a construction site according to claim 1, characterized in that: The execution of the control instruction of the controllable device at the current moment comprises the following steps: Obtain line impedance through impedance measurement device and calculate real-time line loss rate ζ line (t): ζ line (t)=1-cosφ V-I (t); Among them, φ V-I (t) is the voltage and current phase difference at time t; based on Set upper and lower output limits; in, is the minimum output limit of distributed power generation, is the optimal photovoltaic output plan at time t, is the optimal diesel generator output plan at time t, is the maximum output limit of distributed power generation, is the load forecast value at time t; Finally, the control instructions of the controllable device at the current moment are executed through the solid-state relay, and the trip protection is triggered when the upper limit is exceeded.

Images