Zero-carbon harbor district green energy micro-grid system modeling and cooperative scheduling method

Abstract

The invention relates to a modeling and cooperative scheduling method for a zero-carbon harbor district green energy micro-grid system. The cooperative scheduling method based on the consistency of the multiple autonomous energy units is provided for solving the problems that harbor district distributed energy is high in intermittency, ship load impact is large, and multi-energy-flow coupling is complex. The method comprises the following steps: firstly, constructing a system model comprising photovoltaic, wind power, combined cooling heating and power, energy storage and ship load; generating a typical scene set by adopting hybrid uncertainty modeling and adaptive importance sampling; establishing an optimization model and a constraint system which take the minimum daily comprehensive cost as a target and consider the renewable energy utilization rate and marginal carbon emission consistency; a three-level time sequence cooperation stabilizing strategy is designed, and cooperative response of combined cooling heating and power and energy storage is achieved; and solving by adopting an improved stingless bee optimization algorithm to obtain an optimal configuration and scheduling strategy. According to the invention, through distributed coordination and carbon emission consistency regulation and control, the operation economy, renewable energy consumption capability and system low-carbon property of the micro-grid in the harbor district are effectively improved.

Description

Technical Field

[0001] This invention belongs to the field of port area energy system management and smart grid technology, specifically involving a method for modeling and collaborative scheduling of a zero-carbon port area green energy microgrid system. Background Technology

[0002] The construction of zero-carbon port area microgrids dominated by renewable energy has become an inevitable trend. However, such systems exhibit highly decentralized characteristics: the output of distributed power sources such as photovoltaics and wind turbines is highly intermittent, and flexible loads such as all-electric ships and refrigerated containers are widespread. In particular, the impact loads when ships dock pose a severe challenge to the system's power balance. Against this backdrop, "autonomous energy units" with both production and consumption capabilities have changed the traditional energy supply model through point-to-point trading. The unique multi-energy flow coupling requirements of cooling, heating, and electricity in port areas further increase the complexity of system operation.

[0003] Existing energy management methods are ill-equipped to effectively address these challenges. Centralized optimization suffers from drawbacks such as high risk of single-point failures and poor flexibility, making it difficult to adapt to the collaborative operation requirements of highly decentralized autonomous energy units. While traditional distributed optimization methods can achieve distributed decision-making to some extent, they often face problems such as model complexity, poor convergence, and difficulty in guaranteeing global optimum when dealing with multi-energy flow coupling and multi-objective optimization. Therefore, there is an urgent need for a highly efficient distributed scheduling method that can efficiently coordinate various "autonomous energy units" and achieve a comprehensive optimization of economy and low carbon emissions while ensuring the overall optimization performance of the system. Summary of the Invention

[0004] This invention addresses the need for economical and low-carbon coordinated operation of zero-carbon port area microgrids by constructing a distributed coordinated scheduling model based on the consistency of multiple autonomous energy units. By abstracting each distributed unit in the system as an "autonomous energy unit" and using the marginal carbon emission rate as the key coordination indicator, the improved spikeless bee optimization algorithm achieves globally optimal scheduling, effectively improving the operational economy, renewable energy absorption capacity, and system scalability of the port area microgrid. The technical problem solved by this invention is achieved through the following technical solution: A method for modeling and collaborative scheduling of a zero-carbon port area green energy microgrid system includes the following steps: Step 1: Construct a model of a zero-carbon port area distributed green energy microgrid system; Step 2: Generate a typical scenario set based on hybrid uncertainty modeling and adaptive importance sampling; Step 3: Construct a multi-objective optimization scheduling model for the zero-carbon port area microgrid; Step 4: Construct model constraints; Step 5: Design a "three-level time-series coordinated" mitigation strategy and adaptive coordinated mechanism that takes into account ship load impacts; Step 6: A novel metaheuristic algorithm—the improved stingless bee optimization algorithm—is used to solve the optimization model of the port area's integrated energy system to obtain the optimal equipment configuration and energy dispatch strategy that balances economy and low carbon emissions. The specific details of each step are as follows: Step 1: Construct a model of a zero-carbon port area distributed green energy microgrid system; First, a mathematical model of the autonomous energy unit is constructed. As a fundamental component of the port area microgrid, the complex energy flow and conversion relationships within the autonomous energy unit profoundly impact the overall system operation characteristics. To achieve zero-carbon operation, zero-carbon energy units such as photovoltaic and wind power are prioritized, and a combined cooling, heating, and power (CCHP) system is integrated to improve energy efficiency. Simultaneously, energy storage systems and impact loads are modeled to lay the foundation for subsequent coordinated optimization scheduling. The specific mathematical model of the autonomous energy unit is shown below: Each autonomous energy unit contains the following energy flow relationships: (1) In the formula, , , These are the net power for electricity, heat, and cooling, respectively. , , These are power generation, heat generation, and cooling capacity, respectively. , , This refers to the energy storage discharge power; , , Power for energy storage charging; , , These are the power of the electrical, heating, and cooling loads, respectively. Secondly, mathematical models for each type of autonomous energy unit are constructed. Different types of autonomous energy units undertake differentiated functions in the port area microgrid, and their unique operating characteristics play a decisive role in the formulation of the system's zero-carbon energy management strategy. The mathematical models for each type of autonomous energy unit are shown below: (1) Photovoltaic autonomous energy unit As the core zero-carbon energy unit of the system, the experimental data on photovoltaic power generation characteristics were analyzed, and it was found that its output power has a linear relationship with the light intensity. The mathematical model is shown below: (2) In the formula, This indicates the maximum output power of the photovoltaic panel under standard test conditions; Indicates the reference illumination intensity under standard test conditions; express tThe photovoltaic output power is calculated in real time. express t The light intensity is measured in real time at all times; (2) Autonomous Wind Power Energy Unit As the core zero-carbon energy unit of the system, the observation data of wind power generation characteristics are analyzed, and it is found that its output power and wind speed satisfy a piecewise function relationship. The mathematical model is as follows: (3) In the formula, express t The actual output power of the fan at any given time; This indicates the wind speed at which the wind turbine will engage; the wind turbine will not generate electricity when the wind speed is below this value. This indicates the fan's cutoff speed; the fan will stop operating when the speed exceeds this value. This indicates the rated wind speed of the fan, at which the fan outputs its rated power. This indicates the rated power of the fan, which is the maximum output power at or above the rated wind speed. (3) Combined cooling, heating and power autonomous energy unit system As a key energy unit for achieving efficient energy utilization and carbon reduction, the operating data of the port area's combined cooling, heating and power (CCHP) system were fitted to derive the conversion relationship between fuel input and electrical, heat and cooling output. The mathematical model is shown below: (4) In the formula, The power generation capacity of the CCHP system represents the electrical power output of the system. The heat output of the CCHP system represents the heat output recovered and utilized by the system. The cooling capacity of the CCHP system represents the cooling capacity converted through the absorption chiller. Average power generation efficiency represents the average conversion efficiency from fuel to electricity; Heat recovery efficiency represents the proportion of waste heat recovered and utilized. This represents the natural gas consumption of the CCHP system, indicating the amount of fuel consumed per unit time. The coefficient of performance (COP) of an absorption refrigeration system represents the cooling power generated per unit of heat power. (4) Ship Autonomous Energy Unit Model Ship arrivals are the primary source of impact load for the port area microgrid. This model accurately describes its impact on the system by establishing a quantitative relationship between ship arrival behavior and load. t Total electrical load It is composed of the port area's basic load and the additional load from arriving ships, and its mathematical model is shown below: (5) In the formula, For microgrids in t Total electrical load power at any given time; For the Hong Kong area t The baseline load power at any given time is not directly affected by the arrival of ships at port. for t At any given moment, the additional load power caused by the arrival of ships in port; As special mobile autonomous energy units, ships' high-capacity electricity demand during their docking significantly alters the load distribution of the port area's microgrid, impacting the power balance and voltage stability of the regional power grid. The specific expression for ship load is shown below: (6) In the formula, for t At any given moment, the additional load power caused by the arrival of ships in port; for t The number of vessels that are constantly at port and operating; For the first k The load power of loading and unloading equipment such as gantry cranes corresponding to each ship; No. k The charging load power of electric vehicles transported by each ship; For the first k The load capacity of newly added office and auxiliary facilities corresponding to each ship; Gantry crane loads are intermittent and impulsive; electric vehicle charging loads are relatively stable but concentrated; after a ship arrives at port, the power consumption of related dispatch centers and command stations increases. To describe it more accurately, each component load is modeled: (7) In the formula, for t The number of gantry cranes currently performing loading and unloading operations; This represents the average load weight. This represents the average lifting speed; The overall energy efficiency for a complete work cycle. g It is the acceleration due to gravity; For the first k The number of electric vehicles being charged on the ship; For the first n Vehicle charging current; For the first n The car is t The open-circuit voltage at any given time is the battery's state of charge. The function; This refers to the battery's internal resistance. For the first k Fixed power of infrastructure in the ship's related office spaces; for t The number of people working in this area at any given time; Power density per capita; The temperature regulation coefficient for a heating, ventilation, and air conditioning system represents the additional power required to maintain the set temperature for every 1 degree Celsius difference. for t Constant outdoor ambient temperature; Set the indoor temperature; (5) Energy conversion relationship Energy conversion equipment plays a crucial role in achieving coordinated operation of multiple energy flows and improving energy efficiency. Its input-output characteristics directly affect the overall energy efficiency of the system. The mathematical models of the main energy conversion equipment are shown below: (8) In the formula, Total heat production power represents the total heat production power generated by the system through various means; The comprehensive heat recovery efficiency represents the overall efficiency of waste heat recovery during the power generation process. Total power generation represents the total output power of all power generation equipment in the system; Average power generation efficiency represents the average conversion efficiency from fuel to electricity; Total cooling power represents the total cooling power generated by the system. The coefficient of performance (COP) for electric refrigeration represents the energy efficiency ratio of an electric refrigeration unit. The electrical power consumed by the electric chiller; (6) Carbon capture system model As a key negative carbon autonomous energy unit for achieving zero-carbon operation, the carbon capture system captures carbon dioxide generated by traditional energy units such as CCHP through its internal physicochemical processes, directly participating in the coordinated scheduling of carbon flow and power balance in the system. This system is a typical multi-energy-port—electricity, heat, and carbon coupled system; its operating characteristics determine that it is both an energy-transfer adjustable load and a carbon flow regulation unit. The mass of carbon dioxide captured per unit time is proportional to the total amount of carbon dioxide produced by the CCHP system from burning natural gas. This total amount is determined by the natural gas consumption and its inherent carbon emission factor. The carbon capture system captures this portion of carbon dioxide with a certain efficiency through a chemical absorption process, and its mathematical model is shown below: (9) In the formula, The mass of carbon dioxide captured per unit time; This serves as the baseline acquisition efficiency for the system. This refers to the natural gas consumption of the CCHP system. Carbon emission factors of natural gas; The operating intensity control factor represents the internal instructions for adjusting the overall operating load rate of the system; The energy quality control factor represents the internal instructions for regulating the level and quality of energy input into the system. This is the load saturation factor, whose value determines the capture rate as a function of the operating load. The rate at which it increases and approaches the saturation value; The energy input efficiency index reflects the level of energy input. The nonlinear effect on the capture rate, when Diminishing returns occur when... The revenue increases over time; The total equivalent electrical power consumption of a carbon capture system is a function of its operating intensity, as shown in the following expression: (10) In the formula, This represents the total equivalent electrical power consumption of the carbon capture system. The fixed standby power of the carbon capture system is the basic electrical load that must be consumed to keep the system in an operational state. The load power factor characterizes the operating load. The square effect on the energy consumption of rotating equipment such as motors and pumps within the system; The capture and processing power coefficient mainly characterizes the electrical energy consumed in the subsequent processing and compression of a unit mass of carbon dioxide. (7) Dynamic characteristics of energy storage systems Energy storage devices play a crucial role in mitigating zero-carbon energy fluctuations and responding to load shocks. The dynamic changes in their energy state directly affect the system's regulation capability. The dynamic characteristics of energy storage systems are shown below: (11) In the formula, for t+ The state of charge at time 1 represents the proportion of remaining energy in the energy storage device; for t The state of charge at any given moment; Charging efficiency represents the energy conversion efficiency of an energy storage device during the charging process. for t Real-time charging power; The time step represents the time interval between adjacent moments; Discharge efficiency represents the energy conversion efficiency during the discharge process of an energy storage device. for t Discharge power at any given moment; This refers to the rated capacity of the energy storage device. Step 2: Generate a typical scenario set based on hybrid uncertainty modeling and adaptive importance sampling; This invention employs a hybrid uncertainty modeling and adaptive intelligent sampling strategy to generate a set of typical scenarios representing multiple uncertainties in a system more accurately and efficiently. This method comprehensively considers stochastic uncertainties such as load fluctuations and renewable energy output, as well as cognitive uncertainties such as extreme weather and policy adjustments, and significantly improves sampling efficiency. (1) Modeling of mixed uncertainty First, a probabilistic-fuzzy hybrid uncertainty model is constructed to differentiate the uncertainty components of the system: (12) In the formula, This represents the total uncertainty component of the system. The component representing the ship load fluctuation follows a normal distribution. ; The error component for photovoltaic power output prediction follows a Weibull distribution. ; The error components for wind power output prediction follow a normal distribution. ; The electrical load forecast error component follows a Beta distribution. ; The unconventional uncertainty component caused by factors such as extreme weather or policy adjustments is represented by a bounded interval. express, Indicates the maximum possible deviation; (2) Adaptive Importance Sampling To improve sampling efficiency, an adaptive importance sampling method is adopted, which dynamically adjusts the sampling distribution to focus on high-probability or high-risk areas that have a significant impact on the optimization objective. First, an initial scene set is generated based on the prior distribution. According to the scenario s Calculate the importance weight of factors contributing to the objective function, such as the rate of change in operating costs: (13) in accordance with Reconstruct the sampling distribution and perform encrypted sampling in high-weight regions: (14) In the formula, For the scene s The importance weight is determined by the fact that a higher weight indicates a greater impact of the scenario on the optimization objective. In the scene s The daily comprehensive cost of the system; This represents the average daily overall system cost across all scenarios. The set of scenes in the current iteration; The reconstructed sampling probability density function; Represents a mean of The variance is The normal distribution is used in the scene. s sample points Local sampling was conducted in the vicinity; For the scene s The vector of uncertain variables; This is a preset bandwidth parameter used to control the range of local sampling; Repeat the above process until the sampling distribution stabilizes, and finally generate a set of typical scenarios with strong representativeness;

[0005] (3) Extraction of typical scenes First, to reduce computational complexity, cluster analysis is used to extract typical scenarios from a large number of scenarios; Extract statistical features from each scene to construct a feature vector: (15) in: (16) (17) (18) In the formula, For the scene s The feature vector is a three-dimensional column vector used to characterize the statistical features of the scene; This represents the average value for the scene. The standard deviation of the scene; The scene fluctuation range is represented by T; T represents the total number of time periods in the scheduling cycle. For the scene s At any moment t The value of the uncertain variable; Secondly, to extract typical scenarios more accurately, a weighted K-means clustering method is employed. The weights in this method are based on the probability of each scenario occurring; scenarios with higher probabilities have a greater impact on the location of cluster centers during the clustering process, thus ensuring that the extracted typical scenarios more accurately represent high-probability operating states. The weighted clustering objective function is: (19) Cluster center update formula: (20) Finally, taking the scenario corresponding to the cluster center as a typical scenario, its probability is: (twenty one) In the formula, K The number of clusters represents the preset number of categories; For the first k The set of all samples in each cluster; For the sample s The probability of; For the sample s eigenvectors; For the first k Cluster centers of each cluster; For the first k The probability of a typical scenario; Step 3: Construct a multi-objective optimization scheduling model for the zero-carbon port area microgrid, as detailed below: Establish a comprehensive optimization target system that considers economic efficiency, low carbon emissions, and renewable energy consumption: First, take the daily comprehensive cost of the system as the optimization target for model configuration, taking into account the total operating cost of the system, the carbon emission cost of the system, the operating penalty cost of the system, and the renewable energy utilization rate. This invention introduces a marginal carbon emission consistency term for autonomous energy units into the objective function. By driving the marginal carbon emission intensity of each unit to converge, it achieves optimal system carbon efficiency. Marginal carbon emission consistency refers to the fact that in a distributed energy system, each autonomous energy unit has a carbon emission intensity coefficient, i.e., the amount of carbon emissions corresponding to a unit of energy output. Through coordinated optimization, the carbon emission intensity coefficients of each unit at the optimal operating point are made equal, at which point the total system carbon emissions are minimized. This mechanism drives the system towards the optimal carbon efficiency state by penalizing the difference in carbon emission intensity between adjacent units. The specific expression for the objective function is: (twenty two) In the formula, F The total daily cost of the system; The unit carbon emission cost coefficient represents the external cost incurred by the system for each unit mass of carbon dioxide emitted. , , Traditional energy unit k The carbon emission coefficient; In the scene s Next, the k The active power output of a traditional energy unit; M This refers to the number of traditional energy autonomous energy units in the system, specifically those units that generate direct carbon emissions; S This represents the total number of typical scenarios; For the scene s The probability of occurrence; is the total net carbon emissions of the system in the scenario s ; is the carbon emission coefficient of the electric power of the k th traditional energy unit, representing the direct carbon dioxide emissions generated per unit of electric energy generated; is the carbon emission coefficient of the thermal power of the k th traditional energy unit, representing the direct carbon dioxide emissions generated per unit of thermal energy generated; is the carbon emission coefficient of the cooling power of the k th traditional energy unit, representing the direct carbon dioxide emissions generated per unit of cooling capacity generated; is the s scenario, and the k th traditional energy unit's active electric power output; is the <000031​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​ For renewable energy utilization rate; This is the renewable energy utilization rate incentive coefficient, representing the degree of encouragement for the consumption of renewable energy power; For the scene s The actual amount of renewable energy power consumed; For the scene s The maximum available renewable energy power; To ensure consistency in marginal carbon emissions; The marginal carbon emission consistency weighting coefficient is the coefficient that is required to level off the marginal carbon emissions between units in order to approximate the optimal carbon efficiency of the system. In the first i In a distributed control structure of an autonomous energy unit, the set of neighboring units that can directly interact with it; , For the first i Carbon emission coefficient of each autonomous energy unit; For the scene s Next, the j Marginal carbon emissions of an autonomous energy unit; Step 4, construct the model constraints, as follows: (1) Power balance constraint: For each scenario s And every moment t The system must satisfy power balance: (twenty three) In the formula, N This represents the total number of autonomous energy units in the system. For the scene s Lower Autonomous Energy Unit i At any moment t Net power; For the scene s Next moment t Power exchange with the main power grid; For the scene s Next moment t Total electrical load power; For the scene s Next moment t Power loss in the power grid; (2) Thermal power balance constraint (twenty four) In the formula, N This represents the total number of autonomous energy units in the system. For the scene s Lower Autonomous Energy Unit i At any momentt Net heat output power; For the scene s Next moment t Total heat load power; (3) Cooling power balance constraint (25) In the formula, N This represents the total number of autonomous energy units in the system. For the scene s Lower Autonomous Energy Unit i At any moment t Net cooling power; For the scene s Next moment t Total cooling load power; (4) Output constraints of power generation equipment For each autonomous energy unit i Scene s and time t : (26) In the formula, , They are autonomous energy units i Maximum and minimum output; For the scene s Lower Autonomous Energy Unit i At any moment t Net cooling power; (5) Charge and discharge power constraints (27) In the formula, For the scene s Below, including energy storage i Each autonomous energy unit at time t The charging power; For the scene s Below, including energy storage i Each autonomous energy unit at time t The discharge power; For the first i The maximum permissible charging power of energy storage devices in an autonomous energy unit; For the first i The maximum allowable discharge power of energy storage devices in an autonomous energy unit; (6) Charge state constraints (28) In the formula, For the scene s Below, autonomous energy uniti At any moment t The state of charge; Autonomous Energy Unit i Minimum permissible state of charge; Autonomous Energy Unit i Maximum permissible state of charge; (7) Power grid interaction constraints (29) In the formula, For the scene s Next moment t Autonomous Energy Unit i and j Transmission power between; For the scene s Next, at the moment t The exchange power between the port area microgrid and the main power grid; The maximum allowable exchange power of the connection lines between the port area microgrid and the main power grid; (8) Power deviation constraints between scenes (30) In the formula, Autonomous Energy Unit i The maximum power deviation between scenes; For the scene s Lower Autonomous Energy Unit i At any moment t Net power; For another different scenario Next, the i Each autonomous energy unit at time t Net power; (9) Operational constraints of carbon capture systems As a crucial dispatchable flexible electrical load within the system, the carbon capture system's operating power must be strictly limited to the equipment's rated capacity. This constraint is key to ensuring the engineering feasibility of the optimized scheduling scheme, preventing invalid scheduling commands that exceed the equipment's physical limits from appearing in the optimization results. The mathematical expression for its operating constraint is as follows: (31) In the formula, For the scene s Operating power of the carbon capture system; , These are the minimum technical output and maximum permissible operating electrical power of the carbon capture system, respectively. For carbon capture systems, there are two start-up and shutdown state variables used to characterize the operation and shutdown status of the equipment during the scheduling cycle; Step 5: Design a "three-level time-series coordinated" mitigation strategy that takes into account ship load impacts; (1) Quantification of ship load impact First, the invention defines the ship load impact rate, normalizes the absolute ship load fluctuation, and improves the applicability of the method. The calculation formula is as follows: (32) In the formula, for t The ship load impact rate at any given time indicates t The degree of impact of the ship's load on the system's baseline load at any given moment; for t Ship load power at any time This is the system's baseline load power; To ensure that the fluctuation grading threshold matches the system's regulation capability, a comprehensive system regulation coefficient is introduced to reflect the regulation potential of the combined cooling, heating, and power (CCHP) system and energy storage. Its calculation formula is as follows: (33) (34) (35) In the formula, This refers to the overall system adjustment coefficient; , The weights representing the regulation capabilities of the CCHP system and the energy storage system are calculated using the analytic hierarchy process (AHP). , They represent t Adjustability coefficients of the CCHP system and energy storage system at any given time; To maximize the output of the CCHP system; This refers to the rated capacity of the energy storage. Based on the system's comprehensive adjustment coefficient, a dynamic hierarchical threshold can be constructed. , : (36) (37) In the formula, , These are the first-order response coefficient and the second-order response coefficient, respectively. (2) Three-level response mechanism Specifically, it is divided into the following three levels, when At that time, a Level 1 response will be initiated. When a Level II response is initiated, A three-level response should be initiated. The core logic of Level 1 response is to use the "thermoelectric coupling" characteristic of the CCHP system to replace traditional power generation regulation with low-cost cogeneration, thereby reducing system regulation costs. The core of CCHP system regulation is the cross-energy flow impact of fuel input changes on electrical and thermal output. This relationship needs to be quantified using an energy flow coupling matrix—the matrix reflects both the conversion relationship of "fuel input to electrical power" and the conversion relationship of "fuel input to thermal power," accurately describing the multi-energy flow regulation behavior of the CCHP system. The CCHP system energy flow coupling matrix is ​​calculated as follows: (38) In the formula, For electrical power output; For thermal power output; Indicates power generation efficiency; Indicates heat recovery efficiency; This represents the waste heat utilization efficiency; based on the coupling coefficient, the adjustment amount of the CCHP system can be calculated according to the change in fuel input, and the calculation formula is as follows: (39) In the formula, This refers to the adjustment value of the CCHP system; This represents the change in natural gas consumption. This represents the change in waste heat recovery. The secondary response, through the energy storage system, directly smooths out fluctuations via energy throughput, acting as a "power buffer" to quickly respond to moderate power imbalances. A precise control model for energy storage power is established, and its charging and discharging power calculations are shown below: (40) (41) In the formula, for t Real-time charging power of energy storage system; The charging energy storage regulation coefficient represents the regulation ratio allocated to energy storage. This refers to the regulating power already undertaken by the CCHP system; The maximum permissible charging power of the energy storage system; for t Real-time energy storage system discharge power; This is the discharge energy storage regulation coefficient; This refers to the maximum permissible discharge power of the energy storage system. The Level 3 response system coordinates and mitigates fluctuations in an optimal cost-efficiency ratio. Total regulation demand is divided into two parts: one part is handled by the CCHP system through multi-energy flow coordination, and the other part is handled by energy storage. In this strategy, the system comprehensive adjustment coefficient Based on the equipment status calculation at the end of the previous scheduling period, the causality of the threshold determination is ensured. A proportional hybrid transition mechanism is adopted between the various response levels. When the impact rate approaches the threshold, the adjustment tasks of CCHP and energy storage are automatically allocated according to the distance ratio to avoid policy jumps. The baseline load in the impact rate calculation is used for this purpose. Take the average load of the day to ensure normalization stability; Step 6: Regarding the model solution, this invention employs a novel metaheuristic algorithm—an improved stingless bee optimization algorithm—to solve the energy storage optimization configuration model. To further enhance the algorithm's global exploration capability, convergence speed, and solution accuracy, this invention makes two key improvements to the standard stingless bee optimization algorithm: First, a chaotic mapping is introduced in the population initialization stage to enhance the diversity and ergodicity of the initial population; second, a nonlinear reduction factor is used in the temperature reduction strategy to more finely balance the exploration and development behavior of the algorithm in different iteration stages. By simulating the cooperative search behavior of stingless bee populations, such as hiring, observing, and scouting, the population position is iteratively updated. Under the premise of satisfying all operational constraints, the optimal configuration scheme that minimizes the daily comprehensive cost of the system and the corresponding 24-hour system operation strategy are dynamically obtained. The following is the calculation formula for the improved stingless bee optimization algorithm: (1) Population initialization based on chaotic mapping To overcome the problems of uneven population distribution and poor diversity that may result from random initialization, this invention uses a cubic chaotic mapping to generate the initial population. This mapping can produce chaotic sequences with good ergodicity and strong randomness, as shown in the following formula: (42) No. i Only one bee in the first j The initial position on the dimension is generated by the following formula: (43) In the formula, For the first n The values ​​of the chaotic variables after +1 iterations; For the first n The value of the chaotic variable in the next iteration has a range of values. ; The values ​​are chaotic sequence values ​​generated by the cubic chaotic mapping; For the first i Only one bee in the first j The initial position of the dimension; , The first j The lower and upper bounds of a dimension; (2) Brooding room construction stage After initialization, the algorithm evaluates the fitness of each individual and designates the individual with the best fitness as the "brooding mother bee". This stage simulates the behavior of worker bees tightly constructing brood chambers in the central area of ​​the hive (brooding zone), corresponding to the utilization phase of the algorithm, which involves a fine-grained local search around the current optimal solution. In this phase, the algorithm generates new candidate solutions around the "brooding mother bee" (the current optimal solution). This process involves two strategies, using a probabilistic approach... Choose which one to execute; Strategy 1: when At that time, new candidate solutions (brooder room) It is a brood mother bee A bee randomly selected from an elite colony And another randomly selected bee Generated by a combination of factors; its position update formula is as follows: (44) In the formula, It is based on the current temperature and maximum temperature The calculated temperature factor is used to adjust the search step size; It is between random numbers, temperature factor The calculation formula is: (45) Current temperature A non-linear decreasing strategy is adopted for updating to better match the search requirements of the optimization process. The calculation formula is improved as follows: (46) In the formula, For the first m The current temperature at the next iteration; Maximum temperature; This represents the maximum number of iterations. This is a non-linear adjustment factor, a real number greater than 0. When At that time, the temperature drops slowly at first and then rapidly, which is beneficial for later localized development; when At that time, the temperature drops rapidly at first and then slowly, which is beneficial for the initial overall exploration; Strategy 2: when At that time, new candidate solutions From the brood mother bee And two other randomly selected bees , Co-generated; (47) In the formula, It is a random number that follows a standard normal distribution; (3) Honeypot construction stage This phase simulates the behavior of worker bees building honeypots to store nectar away from the center of the hive, corresponding to the algorithm's exploration phase, which involves a broad global search across the entire search space. In this phase, the algorithm randomly selects a bee as the "honeypot queen" and generates new candidate solutions around it. This process also incorporates two strategies, using probability... Choose; Strategy 1: when At that time, new candidate solutions From honeypot queen bees Current bees And another random bee Co-generation: (48) Strategy 2: when At that time, new candidate solutions From honeypot queen bees And two other random bees , Co-generation: (49) New candidate solutions are generated in these two stages. Then, the algorithm will perform a greedy selection. If a new solution is found... Its fitness is better than the current solution. If the current solution is not found, the new solution will replace it. (4) Foraging stage of wax sources This phase simulates the behavior of some worker bees searching for new wax sources when there is insufficient wax in the hive. This corresponds to the mechanism in the algorithm to avoid getting trapped in local optima; the algorithm sets a trial counter for each bee. If the location of a particular bee does not improve after the brooder or honeypot construction phase, its corresponding... It will increase by one. When When the bee exceeds a preset limit, it is considered "idle" and its position will be reinitialized. The reinitialization adopts the chaotic mapping strategy consistent with the population initialization stage, that is, generating a new position according to Equations (42) and (43) to enhance population diversity and help the algorithm escape local optima; The improved spiky bee optimization algorithm described above, which is well known to those skilled in the art, is a modeling and collaborative scheduling method for a zero-carbon port area green energy microgrid system. The advantages and positive effects of this invention are: 1. This invention constructs a distributed collaborative scheduling model based on the consistency of multiple autonomous energy units. By introducing the marginal carbon emission rate as a key coordination variable and using a hybrid uncertainty modeling method to handle multiple uncertainties in the system, it provides a systematic solution for achieving the economic and low-carbon coordinated operation of a zero-carbon port area microgrid. 2. This invention designs a "three-level time-series coordinated" mitigation strategy that takes into account the impact of ship loads. By quantifying the ship load impact rate and establishing a dynamic graded response mechanism, combined with the multi-energy flow coordination of the CCHP system and the power regulation of the energy storage system, the port area microgrid's ability to cope with impact loads and its operational stability are effectively improved. 3. This invention proposes an improved stingless bee optimization algorithm. By introducing chaotic mapping initialization and nonlinear temperature reduction strategies, the algorithm's global exploration and local development capabilities are optimized, providing an efficient and stable computational method for solving high-dimensional, nonlinear port area integrated energy system optimization models. Attached Figure Description

[0006] Figure 1 This is a flowchart of a modeling and collaborative scheduling method for a zero-carbon port area green energy microgrid system. Figure 2 This is a typical daily electricity, heat, and cooling load curve for the port area in this embodiment; Figure 3 This is a comparison chart of the power of key buses before and after the "three-level timing coordination" suppression strategy in this embodiment; Figure 4 This is a diagram showing the system power balance and energy storage operation status in this embodiment; Figure 5 This is a convergence characteristic curve of the improved stingless bee optimization algorithm in this embodiment; Figure 6 This is a comparative analysis chart of carbon emissions in this embodiment; Detailed Implementation

[0007] Figure 1The process of modeling and collaborative scheduling of a zero-carbon port area green energy microgrid system is as follows: Start → Construct a multi-type autonomous energy unit system architecture including photovoltaic, wind power, combined cooling, heating and power (CCHP), energy storage, and ship load, and establish mathematical models for each unit → Generate a typical scenario set representing multiple uncertainties of the system based on mixed uncertainty modeling and adaptive importance sampling method → ​​Construct a multi-objective optimization scheduling model with the goal of minimizing the system's daily comprehensive cost and taking into account the consistency between renewable energy utilization rate and marginal carbon emissions → Establish a complete constraint system including power balance, equipment output, energy storage dynamics, grid interaction, and deviations between scenarios → Design a "three-level time-series collaborative" mitigation strategy based on dynamic classification of ship load impact rate and system comprehensive adjustment coefficient to achieve collaborative response of CCHP and energy storage → Solve the model using an improved stingless bee optimization algorithm, and iteratively obtain the economic and low-carbon optimal solution with energy storage configuration and operation strategy as decision variables → End; Example This invention is a method for modeling and collaborative scheduling of a zero-carbon port area green energy microgrid system, the flowchart of which is shown below. Figure 1 As shown; Step 1: Construct a model of a zero-carbon port area distributed green energy microgrid system; Step 2: Generate a typical scenario set based on hybrid uncertainty modeling and adaptive importance sampling; Step 3: Construct a multi-objective optimization scheduling model for the zero-carbon port area microgrid; Step 4: Construct model constraints; Step 5: Design a "three-level time-series coordinated" mitigation strategy and adaptive coordinated mechanism that takes into account ship load impacts; Step 6: A novel metaheuristic algorithm—the improved stingless bee optimization algorithm—is used to solve the optimization model of the port area's integrated energy system to obtain the optimal equipment configuration and energy dispatch strategy that balances economy and low carbon emissions. Step 1: Construct a distributed green energy microgrid system model for the zero-carbon port area. First, use equation (1) to establish the balance relationship of electrical, thermal, and cold energy flows within the autonomous energy unit; second, use equations (2)-(4) to establish mathematical models for power generation autonomous energy units such as photovoltaic, wind turbine, and combined cooling, heating, and power; then, use equations (5)-(7) to model the load characteristics of the ship as a special mobile autonomous energy unit; finally, use equations (8) and (11) to establish the system energy conversion relationship and the dynamic characteristic model of the energy storage system. Step 2: Generating a typical scenario set based on hybrid uncertainty modeling and adaptive importance sampling. First, a probability-fuzzy hybrid uncertainty model is constructed, and Equation (12) is used to differentiate the uncertainty components of the system. Second, adaptive importance sampling is achieved through Equations (13)-(14), dynamically adjusting the sampling distribution to focus on high-probability or high-risk areas that have a significant impact on the optimization objective. Finally, based on Equations (15)-(21), a weighted K-means clustering method is used to extract typical scenarios from a large number of scenarios to reduce computational complexity. Step 3: Construct a multi-objective optimization scheduling model for the zero-carbon port area microgrid. The function established by equation (22) is used as the optimization objective, which comprehensively considers the system operating cost, carbon emission cost, operating penalty cost, renewable energy utilization rate index, and marginal carbon emission consistency term of autonomous energy units, so as to achieve synergistic optimization of system economy, low carbon and high renewable energy consumption; Step 4, construct model constraints. Systematically establish the operational and safety constraints of the model. Equations (23)-(25) are the power balance constraints for electricity, heat, and cold to ensure the instantaneous balance of supply and demand of multiple energy flows in the system; Equation (26) is the power output constraint for power generation equipment; Equation (27) is the power constraint for energy storage charging and discharging; Equation (28) is the power state constraint for energy storage; Equation (29) is the power constraint for grid interaction; Equation (30) is the power deviation constraint between scenarios to ensure the robustness of the system operation scheme. Step 5: Design a three-level time-series coordinated mitigation strategy and adaptive coordination mechanism for ship load impact. First, define the ship load impact rate based on equation (32) and quantify load fluctuations. Second, calculate the system comprehensive adjustment coefficient and construct a dynamic graded threshold using equations (33)-(37). Then, determine the response level to be executed based on the threshold range of the impact rate. Finally, achieve economical and efficient mitigation of ship load impact through the coordination of multi-energy flow of the CCHP system and power regulation of the energy storage system based on equations (38)-(41). Step 6: The improved stingless bee optimization algorithm is used to solve the optimization model of the port area integrated energy system. First, the population initialization based on chaotic mapping is performed using equations (42)-(43); second, the brooding chamber construction and honeypot construction stages of the algorithm are executed through equations (44)-(49), respectively, to carry out local development and global exploration; finally, the wax source foraging stage mechanism is used to avoid getting trapped in local optima, thereby effectively improving the global optimization performance and convergence stability of the algorithm, and finally obtaining the optimal equipment configuration and energy scheduling strategy that takes into account both economy and low carbon emissions. This invention uses the operational data of a typical day throughout the year in a port area and the main parameters listed in Table 1 to analyze and evaluate the method.

[0008] Table 1 Main Parameter Settings

[0009] Figure 2 These are the electricity, heat, and cooling load curves for a typical day in the port area. The electricity load curve exhibits a high daytime and low nighttime characteristic, and includes peak loads caused by ship operations. The heat and cooling load curves reflect heating and cooling demands, respectively, and their shapes are related to the energy consumption characteristics of the port area and the ambient temperature. These three curves together constitute the basic load data for the optimized scheduling of the multi-energy flow system, providing input for subsequent modeling and simulation. Figure 3 The changes in critical bus power before and after the implementation of a three-level time-series coordinated mitigation strategy were compared. The original power curve experienced severe fluctuations and peak impacts due to ship operations. After strategy adjustment, the power fluctuations were effectively suppressed, and the curve became smoother. This result verifies the practical effectiveness of the three-level response mechanism in responding to load shocks and maintaining system power stability. Figure 4 The system power balance and energy storage operation status are illustrated. The system power balance diagram reflects the real-time matching relationship between power generation, load, and exchange power. The energy storage operation status diagram shows that the energy storage is in a charging state when renewable energy output is high, and switches to discharging during peak load periods or ship impact periods, indicating that the energy storage plays a role in energy buffering and power support in the system. Figure 5 To improve the convergence characteristic curve of the stingless bee optimization algorithm, the horizontal axis represents the number of iterations, and the vertical axis represents the fitness value. The curve shows that the algorithm gradually approaches the optimal solution with each iteration, decreasing rapidly in the early stages and then stabilizing. This convergence process demonstrates that the algorithm possesses effective optimization ability and stability in solving this optimization problem. Figure 6 The system carbon emissions were compared under three scenarios. Scenario 1 uses traditional centralized scheduling, Scenario 2 employs distributed collaborative scheduling based on marginal carbon emission consistency, and Scenario 3 further integrates a three-level collaborative strategy and an improved algorithm. Chart data shows that carbon emissions gradually decrease from Scenario 1 to Scenario 3. This result demonstrates the practical effectiveness of this invention in reducing system carbon emissions. The port area microgrid system is evaluated based on the objective function. Lower daily comprehensive costs indicate better system economic efficiency, while higher renewable energy utilization rates indicate superior low-carbon performance. Three scenarios are set up, and comparative experiments are conducted to verify the effectiveness of the proposed model and method: Scenario 1: Scheduling is performed using traditional centralized optimization methods; Scenario 2: Employing the distributed collaborative scheduling method based on marginal carbon emission consistency described in this invention; Scenario 3: Based on Scenario 2, the "three-level temporal coordination" smoothing strategy described in this invention and the improved stingless bee optimization algorithm are further integrated for solving the problem; As shown in Table 2, the system operation results are presented for each scenario. In Scenario 1, the total system cost is high due to the inability to fully coordinate the autonomy of the "autonomous energy units." Scenarios 2 and 3, by adopting distributed collaborative scheduling and based on their decentralized autonomous architecture, achieve global economic optimization at the system level, improving operational economy. Scenario 3 further obtains the optimal configuration and operation scheme through advanced smoothing strategies and optimization algorithms, achieving the best balance between economy and low carbon emissions.

[0010] Table 2 System optimization results under different cases

[0011] The results are analyzed as follows: Scenario 1: Due to the rigidity of the optimization model, it is difficult to effectively coordinate the autonomous operation of each unit, resulting in the worst economic efficiency of the scheduling strategy, the lowest utilization rate of renewable energy, and the highest carbon emissions; Scenario 2: By using a distributed collaborative algorithm based on marginal carbon emission consistency, system-level collaborative optimization is achieved. As a result, the daily comprehensive cost and carbon emissions are significantly lower than in Scenario 1, and the utilization rate of renewable energy is effectively improved. Scenario 3: Building upon Scenario 2, the "three-level temporal coordination" mitigation strategy and improved stingless bee optimization algorithm of this invention were fully applied. Results show that this complete solution not only fully leverages the optimization potential of distributed coordination but also achieves optimal results in terms of economy, low carbon emissions, and renewable energy consumption through targeted strategies and efficient algorithms, fully validating the advanced nature and necessity of the overall technical solution of this invention.

[0012] It should be emphasized that the embodiments described in this invention are illustrative rather than limiting. Therefore, this invention includes, but is not limited to, the embodiments described in the specific implementation. Any other implementations derived by those skilled in the art based on the technical solutions of this invention are also within the scope of protection of this invention.

Claims

1. A method for modeling and collaborative scheduling of a zero-carbon port area green energy microgrid system, characterized in that... Includes the following steps: Step 1: Construct a model of a zero-carbon port area distributed green energy microgrid system; Step 2: Generate a typical scenario set based on hybrid uncertainty modeling and adaptive importance sampling; Step 3: Construct a multi-objective optimization scheduling model for the zero-carbon port area microgrid; Step 4: Construct model constraints; Step 5: Design a "three-level time-series coordinated" mitigation strategy and adaptive coordinated mechanism that takes into account ship load impacts; Step 6: A novel metaheuristic algorithm—the improved stingless bee optimization algorithm—is used to solve the optimization model of the port area's integrated energy system to obtain the optimal equipment configuration and energy dispatch strategy that balances economy and low carbon emissions. The specific details of each step are as follows: Step 1: Construct a model of a zero-carbon port area distributed green energy microgrid system; First, a mathematical model of the autonomous energy unit is constructed. As the basic component of the port area microgrid, the complex energy flow and conversion relationships within the autonomous energy unit have a profound impact on the overall operating characteristics of the system. To achieve zero-carbon operation of the system, zero-carbon energy units such as photovoltaic and wind power are prioritized, and a combined cooling, heating and power system is integrated to improve energy efficiency. At the same time, the energy storage system and impact loads are modeled to lay the foundation for subsequent collaborative optimization scheduling. The specific mathematical model of the autonomous energy unit is shown below: Each autonomous energy unit contains the following energy flow relationships: (1) In the formula, , , These are the net power for electricity, heat, and cooling, respectively. , , These are power generation, heat generation, and cooling capacity, respectively. , , This refers to the energy storage discharge power; , , Power for energy storage charging; , , These are the power of the electrical, heating, and cooling loads, respectively. Secondly, mathematical models for each type of autonomous energy unit are constructed. Different types of autonomous energy units undertake differentiated functions in the port area microgrid, and their unique operating characteristics play a decisive role in the formulation of the system's zero-carbon energy management strategy. The mathematical models for each type of autonomous energy unit are shown below: (1) Photovoltaic autonomous energy unit As the core zero-carbon energy unit of the system, the experimental data on photovoltaic power generation characteristics were analyzed, and it was found that its output power has a linear relationship with the light intensity. The mathematical model is shown below: (2) In the formula, This indicates the maximum output power of the photovoltaic panel under standard test conditions; Indicates the reference illumination intensity under standard test conditions; express t The photovoltaic output power is calculated in real time. express t The light intensity is measured in real time at all times; (2) Autonomous Wind Power Energy Unit As the core zero-carbon energy unit of the system, the observation data of wind power generation characteristics are analyzed, and it is found that its output power and wind speed satisfy a piecewise function relationship. The mathematical model is as follows: (3) In the formula, express t The actual output power of the fan at any given time; This indicates the wind speed at which the wind turbine will engage; the wind turbine will not generate electricity when the wind speed is below this value. This indicates the fan's cutoff speed; the fan will stop operating when the speed exceeds this value. This indicates the rated wind speed of the fan, at which the fan outputs its rated power. This indicates the rated power of the fan, which is the maximum output power at or above the rated wind speed. (3) Combined cooling, heating and power autonomous energy unit system As a key energy unit for achieving efficient energy utilization and carbon reduction, the operating data of the port area's combined cooling, heating and power (CCHP) system were fitted to derive the conversion relationship between fuel input and electrical, heat and cooling output. The mathematical model is shown below: (4) In the formula, The power output of the CCHP system represents the electrical power output of the system. The heat output of the CCHP system represents the heat output recovered and utilized by the system. The cooling capacity of the CCHP system represents the cooling capacity converted through the absorption chiller. Average power generation efficiency represents the average conversion efficiency from fuel to electricity; Heat recovery efficiency represents the proportion of waste heat recovered and utilized. This represents the natural gas consumption of the CCHP system, indicating the amount of fuel consumed per unit time. The coefficient of performance (COP) of an absorption refrigeration system represents the cooling power generated per unit of heat power. (4) Ship Autonomous Energy Unit Model Ship arrivals are the primary source of impact load for port microgrids. This model accurately describes their impact on the system by establishing a quantitative relationship between ship arrival behavior and load. t Total electrical load It is composed of the port area's basic load and the additional load from arriving ships, and its mathematical model is shown below: (5) In the formula, For microgrids in t Total electrical load power at any given time; For the Hong Kong area t The baseline load power at any given time is not directly affected by the arrival of ships at port. for t At any given moment, the additional load power caused by the arrival of ships in port; As special mobile autonomous energy units, ships' high-capacity electricity demand during their docking significantly alters the load distribution of the port area's microgrid, impacting the power balance and voltage stability of the regional power grid. The specific expression for ship load is shown below: (6) In the formula, for t At any given moment, the additional load power caused by the arrival of ships in port; for t The number of vessels that are constantly at port and engaged in operations; For the first k The load power of loading and unloading equipment such as gantry cranes corresponding to each ship; No. k The charging load power of electric vehicles transported by each ship. For the first k The load capacity of newly added office and auxiliary facilities corresponding to each ship; Gantry crane loads are intermittent and impulsive; electric vehicle charging loads are relatively stable but concentrated; after a ship arrives at port, the power consumption of related dispatch centers and command stations increases. To describe it more accurately, each component load is modeled: (7) In the formula, for t The number of gantry cranes currently performing loading and unloading operations; This represents the average load weight. This represents the average lifting speed; The overall energy efficiency for a complete work cycle. g It is the acceleration due to gravity; For the first k The number of electric vehicles being charged on the ship; For the first n Vehicle charging current; For the first n The car is t The open-circuit voltage at any given time is the battery's state of charge. The function; This refers to the battery's internal resistance. For the first k Fixed power of infrastructure in the ship's related office spaces; for t The number of people working in this area at any given time; Power density per capita; The temperature regulation coefficient for a heating, ventilation, and air conditioning system represents the additional power required to maintain the set temperature for every 1 degree Celsius difference. for t Constant outdoor ambient temperature; Set the indoor temperature; (5) Energy conversion relationship Energy conversion equipment plays a crucial role in achieving coordinated operation of multiple energy flows and improving energy efficiency. Its input-output characteristics directly affect the overall energy efficiency of the system. The mathematical models of the main energy conversion equipment are shown below: (8) In the formula, Total heat production power represents the total heat production power generated by the system through various means; The comprehensive heat recovery efficiency represents the overall efficiency of waste heat recovery during the power generation process. Total power generation represents the total output power of all power generation equipment in the system; Average power generation efficiency represents the average conversion efficiency from fuel to electricity; Total cooling power represents the total cooling power generated by the system. The coefficient of performance (COP) for electric refrigeration represents the energy efficiency ratio of an electric refrigeration unit. The electrical power consumed by the electric chiller; (6) Carbon capture system model As a key negative carbon autonomous energy unit for achieving zero-carbon operation, the carbon capture system captures carbon dioxide generated by traditional energy units such as CCHP through its internal physicochemical processes, and directly participates in the coordinated scheduling of carbon flow and power balance of the system. The system is a typical multi-energy port - an electrical, thermal and carbon coupled system. Its operating characteristics determine that it is both an energy transfer type adjustable load and a carbon flow regulation unit. The mass of carbon dioxide captured per unit time is proportional to the total amount of carbon dioxide produced by the CCHP system burning natural gas. This total amount is determined by natural gas consumption and its inherent carbon emission factor; the carbon capture system captures this portion of carbon dioxide with a certain efficiency through a chemical absorption process, and its mathematical model is shown below: (9) In the formula, The mass of carbon dioxide captured per unit time; This serves as the baseline acquisition efficiency for the system. This refers to the natural gas consumption of the CCHP system. Carbon emission factors of natural gas; The operating intensity control factor represents an internal instruction for adjusting the overall operating load rate of the system. The energy quality control factor represents the internal instructions for regulating the level and quality of energy input into the system. This is the load saturation factor, whose value determines the capture rate as a function of the operating load. The rate at which it increases and approaches the saturation value; The energy input efficiency index reflects the level of energy input. The nonlinear effect on the capture rate, when Diminishing returns occur when... The revenue increases over time; The total equivalent electrical power consumption of a carbon capture system is a function of its operating intensity, as shown in the following expression: (10) In the formula, This represents the total equivalent electrical power consumption of the carbon capture system. The fixed standby power of the carbon capture system is the basic electrical load that must be consumed to keep the system in an operational state. The load power factor characterizes the operating load. The square effect on the energy consumption of rotating equipment such as motors and pumps within the system; The capture and processing power coefficient mainly characterizes the electrical energy consumed in the subsequent processing and compression of a unit mass of carbon dioxide. (7) Dynamic characteristics of energy storage systems Energy storage devices play a crucial role in mitigating zero-carbon energy fluctuations and responding to load shocks. The dynamic changes in their energy state directly affect the system's regulation capability. The dynamic characteristics of energy storage systems are shown below: (11) In the formula, for t+ The state of charge at time 1 represents the proportion of remaining energy in the energy storage device; for t The state of charge at any given moment; Charging efficiency represents the energy conversion efficiency of an energy storage device during the charging process. for t Real-time charging power; The time step represents the time interval between adjacent moments; Discharge efficiency represents the energy conversion efficiency during the discharge process of an energy storage device. for t Discharge power at any given moment; This refers to the rated capacity of the energy storage device. Step 2: Generate a typical scenario set based on hybrid uncertainty modeling and adaptive importance sampling; This invention employs a hybrid uncertainty modeling and adaptive intelligent sampling strategy to generate a set of typical scenarios that represent multiple uncertainties in a system more accurately and efficiently. This method comprehensively considers random uncertainties such as load fluctuations and renewable energy output, as well as cognitive uncertainties such as extreme weather and policy adjustments, and significantly improves sampling efficiency. (1) Modeling of mixed uncertainty First, a probabilistic-fuzzy hybrid uncertainty model is constructed to differentiate the uncertainty components of the system: (12) In the formula, This represents the total uncertainty component of the system. The component representing the ship load fluctuation follows a normal distribution. ; The error component for photovoltaic power output prediction follows a Weibull distribution. ; The error components for wind power output prediction follow a normal distribution. ; The electrical load forecast error component follows a Beta distribution. ; The unconventional uncertainty component caused by factors such as extreme weather or policy adjustments is represented by a bounded interval. express, Indicates the maximum possible deviation; (2) Adaptive Importance Sampling To improve sampling efficiency, an adaptive importance sampling method is adopted, which dynamically adjusts the sampling distribution to focus on high-probability or high-risk areas that have a significant impact on the optimization objective. First, an initial scene set is generated based on the prior distribution. According to the scenario s Calculate the importance weight of factors contributing to the objective function, such as the rate of change in operating costs: (13) in accordance with Reconstruct the sampling distribution and perform encrypted sampling in high-weight regions: (14) In the formula, For the scene s The importance weight is determined by the fact that a higher weight indicates a greater impact of the scenario on the optimization objective. In the scene s The daily comprehensive cost of the system; This represents the average daily overall system cost across all scenarios. The set of scenes in the current iteration; The reconstructed sampling probability density function; To represent a mean is The variance is The normal distribution is used in the scene. s sample points Local sampling was conducted in the vicinity; For the scene s The vector of uncertain variables; This is a preset bandwidth parameter used to control the range of local sampling; Repeat the above process until the sampling distribution is stable, and finally generate a set of typical scenes with strong representativeness; (3) Extraction of typical scenes First, to reduce computational complexity, cluster analysis is used to extract typical scenarios from a large number of scenarios; Extract statistical features from each scene to construct a feature vector: (15) in: (16) (17) (18) In the formula, For the scene s The feature vector is a three-dimensional column vector used to characterize the statistical features of the scene; This represents the average value for the scene. The standard deviation of the scene; The scene fluctuation range is represented by T; T represents the total number of time periods in the scheduling cycle. For the scene s At any moment t The value of the uncertain variable; Secondly, to extract typical scenarios more accurately, a weighted K-means clustering method is adopted. The weights of this method are based on the probability of occurrence of each scenario. Scenarios with higher probabilities have a greater impact on the location of the cluster centers during the clustering process, thus ensuring that the extracted typical scenarios can more accurately represent high-probability operating states. The objective function of the weighted clustering is: (19) Cluster center update formula: (20) Finally, taking the scenario corresponding to the cluster center as a typical scenario, its probability is: (21) In the formula, K The number of clusters represents the preset number of categories; For the first k The set of all samples in each cluster; For the sample s The probability of; For the sample s eigenvectors; For the first k Cluster centers of each cluster; For the first k The probability of a typical scenario; Step 3: Construct a multi-objective optimization scheduling model for the zero-carbon port area microgrid, as detailed below: Establish a comprehensive optimization target system that considers economic efficiency, low carbon emissions, and renewable energy consumption: First, take the daily comprehensive cost of the system as the optimization target for model configuration, taking into account the total operating cost of the system, the carbon emission cost of the system, the operating penalty cost of the system, and the renewable energy utilization rate. This invention introduces a marginal carbon emission consistency term for autonomous energy units into the objective function. By driving the marginal carbon emission intensity of each unit to tend towards consistency, the system achieves optimal carbon efficiency. Marginal carbon emission consistency means that in a distributed energy system, each autonomous energy unit has a carbon emission intensity coefficient, which is the amount of carbon emissions corresponding to a unit of energy output. Through coordinated optimization, the carbon emission intensity coefficients of each unit at the optimal operating point tend to be equal, at which point the total carbon emissions of the system are minimized. This mechanism drives the system to approach the optimal carbon efficiency state by penalizing the difference in carbon emission intensity between adjacent units. The specific expression for the objective function is: (22) Wherein, F is the daily comprehensive cost of the system; is the unit carbon emission cost coefficient, representing the external cost borne for each unit mass of carbon dioxide emitted in the system; , , are the traditional energy units k 's carbon emission coefficient; is under the scenario s , the k th active power output of the traditional energy unit; M is the number of traditional energy autonomous energy units in the system, specifically referring to the units that will generate direct carbon emissions; S is the total number of typical scenarios; is the probability of occurrence of the scenario s ; is the total net carbon emissions of the system under the scenario s ; is the k th electric power carbon emission coefficient of the traditional energy unit, indicating the direct carbon dioxide emission generated per unit of electric energy generated; is the k th thermal power carbon emission coefficient of the traditional energy unit, indicating the direct carbon dioxide emission generated per unit of thermal energy generated; is the k th cooling power carbon emission coefficient of the traditional energy unit, indicating the direct carbon dioxide emission generated per unit of cooling capacity generated; is under the scenario s , the k th active electric power output of the traditional energy unit; ​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​The active power output of each autonomous energy unit; The unit cost coefficient for power interaction with the main power grid; For the scene s The exchange power between the microgrid and the main grid; Penalty costs for system operation; , , These are the penalty coefficients for wind and solar power curtailment, load reduction, and power imbalance, respectively. For the scene s The amount of wind and solar power that has been curtailed is the amount of renewable energy that has not been utilized. For the scene s The load reduction amount, i.e. the unmet rigid load; For the scene s The power imbalance is used to measure the system safety deviation. For renewable energy utilization rate; This is the renewable energy utilization rate incentive coefficient, representing the degree of encouragement for the consumption of renewable energy power; For the scene s The actual amount of renewable energy power consumed; For the scene s The maximum power of available renewable energy sources; To ensure consistency in marginal carbon emissions; The marginal carbon emission consistency weighting coefficient is the coefficient that indicates the need to level off the marginal carbon emissions between units in order to approximate the optimal carbon efficiency of the system. In the first i In a distributed control structure of an autonomous energy unit, the set of neighboring units that can directly interact with it; , For the first i Carbon emission coefficient of each autonomous energy unit; For the scene s Next, the j Marginal carbon emissions of an autonomous energy unit; Step 4, construct the model constraints, as follows: (1) Power balance constraint: For each scenario s And every moment t The system must satisfy power balance: (23) In the formula, N This represents the total number of autonomous energy units in the system. For the scene s Lower Autonomous Energy Unit i At any moment t Net power; For the scene s Next moment t Power exchange with the main power grid; For the scene s Next moment t Total electrical load power; For the scene s Next moment t Power loss in the power grid; (2) Thermal power balance constraint (24) In the formula, N This represents the total number of autonomous energy units in the system. For the scene s Lower Autonomous Energy Unit i At any moment t Net heat output power; For the scene s Next moment t Total heat load power; (3) Cooling power balance constraint (25) In the formula, N This represents the total number of autonomous energy units in the system. For the scene s Lower Autonomous Energy Unit i At any moment t Net cooling power; For the scene s Next moment t Total cooling load power; (4) Output constraints of power generation equipment For each autonomous energy unit i Scene s and time t : (26) In the formula, , They are autonomous energy units i Maximum and minimum output; For the scene s Lower Autonomous Energy Unit i At any moment t Net cooling power; (5) Charge and discharge power constraints (27) In the formula, For the scene s Below, including energy storage i Each autonomous energy unit at time t The charging power; For the scene s Below, including energy storage i Each autonomous energy unit at time t The discharge power; For the first i The maximum permissible charging power of energy storage devices in an autonomous energy unit; For the first i The maximum allowable discharge power of energy storage devices in an autonomous energy unit; (6) Charge state constraints (28) In the formula, For the scene s Below, autonomous energy unit i At any moment t The state of charge; Autonomous Energy Unit i Minimum permissible state of charge; Autonomous Energy Unit i Maximum permissible state of charge; (7) Power grid interaction constraints (29) In the formula, For the scene s Next moment t Autonomous Energy Unit i and j Transmission power between; For the scene s Next, at the moment t The exchange power between the port area microgrid and the main power grid; The maximum allowable exchange power of the connection lines between the port area microgrid and the main power grid; (8) Power deviation constraints between scenes (30) In the formula, Autonomous Energy Unit i The maximum power deviation between scenes; For the scene s Lower Autonomous Energy Unit i At any moment t Net power; For another different scenario Next, the i Each autonomous energy unit at time t Net power; (9) Operational constraints of carbon capture systems As a crucial dispatchable flexible electrical load within the system, the carbon capture system's operating power must be strictly limited to the equipment's rated operating capacity. This constraint is key to ensuring the engineering feasibility of the optimized scheduling scheme and preventing invalid scheduling instructions that exceed the equipment's physical limits from appearing in the optimization results. The mathematical expression for its operating constraint is as follows: (31) In the formula, For the scene s Operating power of the carbon capture system; , These are the minimum technical output and maximum permissible operating electrical power of the carbon capture system, respectively. For carbon capture systems, there are two start-up and shutdown state variables used to characterize the operation and shutdown status of the equipment during the scheduling cycle; Step 5: Design a "three-level time-series coordinated" mitigation strategy that takes into account ship load impacts; (1) Quantification of ship load impact First, the invention defines the ship load impact rate, normalizes the absolute ship load fluctuation, and improves the applicability of the method. The calculation formula is as follows: (32) In the formula, for t The ship load impact rate at any given time indicates t The degree of impact of the ship's load on the system's baseline load at any given moment; for t Ship load power at any time This is the system's baseline load power; To ensure that the fluctuation grading threshold matches the system's regulation capability, a comprehensive system regulation coefficient is introduced to reflect the regulation potential of the combined cooling, heating, and power (CCHP) system and energy storage. Its calculation formula is as follows: (33) (34) (35) In the formula, This is the overall system adjustment coefficient; , The weights representing the regulation capabilities of the CCHP system and the energy storage system are calculated using the analytic hierarchy process (AHP). , They represent t Adjustability coefficients of the CCHP system and energy storage system at any given time; To maximize the output of the CCHP system; This refers to the rated capacity of the energy storage. Based on the system's comprehensive adjustment coefficient, a dynamic hierarchical threshold can be constructed. , : (36) (37) In the formula, , These are the first-order response coefficient and the second-order response coefficient, respectively. (2) Three-level response mechanism Specifically, it is divided into the following three levels, when At that time, a Level 1 response will be initiated. When a Level II response is initiated, A three-level response should be initiated. The core logic of Level 1 response is to use the "thermoelectric coupling" characteristic of the CCHP system to replace traditional power generation regulation with low-cost cogeneration, thereby reducing system regulation costs. The core of CCHP system regulation is the cross-energy flow impact of fuel input changes on electrical and thermal output. This correlation needs to be quantified through an energy flow coupling matrix—the matrix reflects both the conversion relationship of "fuel input to electrical power" and the conversion relationship of "fuel input to thermal power," accurately describing the multi-energy flow regulation behavior of the CCHP system. The CCHP system energy flow coupling matrix is ​​calculated as follows: (38) In the formula, For electrical power output; For thermal power output; Indicates power generation efficiency; Indicates heat recovery efficiency; This represents the waste heat utilization efficiency; based on the coupling coefficient, the adjustment amount of the CCHP system can be calculated according to the change in fuel input, and the calculation formula is as follows: (39) In the formula, This refers to the adjustment value of the CCHP system; This represents the change in natural gas consumption. This represents the change in waste heat recovery. The secondary response directly smooths out fluctuations through energy throughput via the energy storage system, acting as a "power buffer" to quickly respond to moderate power imbalances. A precise control model for energy storage power is established, and its charging and discharging power is calculated as follows: (40) (41) In the formula, for t Real-time charging power of energy storage system; The charging energy storage regulation coefficient represents the regulation ratio allocated to energy storage. This refers to the regulating power already undertaken by the CCHP system; The maximum permissible charging power of the energy storage system; for t Real-time energy storage system discharge power; This is the discharge energy storage regulation coefficient; This refers to the maximum permissible discharge power of the energy storage system. The Level 3 response coordinates and smooths out fluctuations in the optimal "cost-efficiency" ratio, dividing the total regulation demand into two parts: one part is undertaken by the CCHP system through multi-energy flow coordination, and the other part is undertaken by energy storage. In this strategy, the system comprehensive adjustment coefficient Based on the equipment status calculation at the end of the previous scheduling period, the causality of the threshold determination is ensured. A proportional hybrid transition mechanism is adopted between the various response levels. When the impact rate approaches the threshold, the adjustment tasks of CCHP and energy storage are automatically allocated according to the distance ratio to avoid policy jumps. The baseline load in the impact rate calculation is used. Take the average load of the day to ensure normalization stability; Step 6: Regarding the solution of the model, this invention employs a novel metaheuristic algorithm—an improved stingless bee optimization algorithm—to solve the energy storage optimization configuration model. To further enhance the algorithm's global exploration capability, convergence speed, and solution accuracy, this invention makes two key improvements to the standard stingless bee optimization algorithm: First, a chaotic mapping is introduced in the population initialization stage to enhance the diversity and ergodicity of the initial population; Second, a nonlinear reduction factor is used in the temperature reduction strategy to more finely balance the exploration and development behavior of the algorithm in different iteration stages. By simulating the cooperative search behavior of the stingless bee population, such as hiring, observing, and scouting, the population position is iteratively updated. Under the premise of satisfying all operational constraints, the optimal configuration scheme that minimizes the daily comprehensive cost of the system and the corresponding 24-hour system operation strategy are dynamically obtained. The following is the calculation formula for the improved stingless bee optimization algorithm: (1) Population initialization based on chaotic mapping To overcome the problems of uneven population distribution and poor diversity that may result from random initialization, this invention uses a cubic chaotic mapping to generate the initial population. This mapping can produce chaotic sequences with good ergodicity and strong randomness, as shown in the following formula: (42) No. i Only one bee in the first j The initial position on the dimension is generated by the following formula: (43) In the formula, For the first n The values ​​of the chaotic variables after +1 iterations; For the first n The value of the chaotic variable in the next iteration has a range of values. ; The values ​​are chaotic sequence values ​​generated by the cubic chaotic mapping; For the first i Only one bee in the first j The initial position of the dimension; , The first j The lower and upper bounds of a dimension; (2) Brooding room construction stage After initialization, the algorithm evaluates the fitness of each individual and designates the individual with the best fitness as the "brood mother bee". This stage simulates the behavior of worker bees tightly constructing brood chambers in the central area of ​​the hive (brooding zone), corresponding to the utilization stage of the algorithm, which involves a fine-grained local search around the current optimal solution. In this stage, the algorithm generates new candidate solutions around the "brooding queen bee" (the current optimal solution). This process includes two strategies, using a probabilistic approach... Choose which one to execute; Strategy 1: when At that time, new candidate solutions (brooder room) It is a brood mother bee A bee randomly selected from an elite colony And another randomly selected bee Generated by a combination of factors; its position update formula is as follows: (44) In the formula, It is based on the current temperature and maximum temperature The calculated temperature factor is used to adjust the search step size; It is between random numbers, temperature factor The calculation formula is: (45) Current temperature A non-linear decreasing strategy is adopted for updating to better match the search requirements of the optimization process. The calculation formula is improved as follows: (46) In the formula, For the first m The current temperature at the next iteration; Maximum temperature; This represents the maximum number of iterations. , is a non-linear adjustment factor, a real number greater than 0, when At that time, the temperature drops slowly at first and then rapidly, which is beneficial for later localized development; when At that time, the temperature drops rapidly at first and then slowly, which is beneficial for the initial overall exploration; Strategy 2: when At that time, new candidate solutions From the brood mother bee And two other randomly selected bees , Co-generated; (47) In the formula, It is a random number that follows a standard normal distribution; (3) Honeypot construction stage This phase simulates the behavior of worker bees building honeypots to store nectar away from the center of the hive, corresponding to the algorithm's exploration phase, which involves a broad global search across the entire search space. In this phase, the algorithm randomly selects a bee as the "honeypot queen" and generates new candidate solutions around it. This process also incorporates two strategies: probabilistic... Choose; Strategy 1: when At that time, new candidate solutions From honeypot queen bees Current bees And another random bee Co-generation: (48) Strategy 2: when At that time, new candidate solutions From honeypot queen bees And two other random bees , Co-generation: (49) New candidate solutions are generated in these two stages. Then, the algorithm will perform a greedy selection; if a new solution is found... Its fitness is better than the current solution. If the current solution is not found, the new solution will replace it. (4) Foraging stage of wax sources This phase simulates the behavior of some worker bees searching for new wax sources when there is insufficient wax in the hive. This corresponds to the mechanism in the algorithm to avoid getting trapped in local optima; the algorithm sets a trial counter for each bee. If the location of a bee does not improve after the brooder or honeypot construction stage, its corresponding It will add one, when When the bee exceeds a preset limit, it is considered "idle" and its position will be reinitialized. The reinitialization adopts the chaotic mapping strategy consistent with the population initialization stage, that is, generating a new position according to Equation (42) and Equation (43) to enhance population diversity and help the algorithm escape local optima.

2. The method for modeling and collaborative scheduling of a zero-carbon port area green energy microgrid system according to claim 1, characterized in that... The optimization solution and hierarchical management strategy are implemented through the following steps: 1) Construct a dynamic graded mitigation strategy. By quantifying the ship load impact rate, calculating the system's comprehensive adjustment coefficient and setting dynamic graded thresholds, a "three-level time-series coordinated" response mechanism is activated based on the differences in impact severity. 2) Design a multi-energy flow coordinated response mechanism. Based on the dynamic hierarchical results, the mechanism achieves economical and efficient mitigation of ship impact loads by adjusting the energy flow coupling matrix of the combined cooling, heating and power system, precisely controlling the power of the energy storage system, and optimizing the cost synergy between the two. 3) An improved stingless bee optimization algorithm is used to solve the problem. By introducing cubic chaotic mapping to initialize the population and a nonlinear temperature reduction strategy, the algorithm's global exploration and local exploitation capabilities are optimized. The optimal configuration and scheduling scheme that satisfies all constraints is searched by utilizing the population update behavior of brood chamber construction, honey pot construction and wax source foraging.

Images