Multi-energy-flow scheduling method and system of electricity-hydrogen-heat low-carbon building energy system

Abstract

The invention discloses a multi-energy-flow scheduling method and system for an electricity-hydrogen-heat low-carbon building energy system. The scheduling method comprises the following steps: S1, establishing a physical characteristic model and an operation constraint condition of each piece of related equipment in the electricity-hydrogen-heat low-carbon building energy system; s2, establishing a target function by taking balance of system operation cost and sustainable operation cost as a target; s3, constructing a Markov decision process model based on the objective function and in combination with the physical feature model and the operation constraint condition of each related device; and S4, performing optimal scheduling on multiple energy flows of the electricity-hydrogen-heat low-carbon building energy system based on the Markov decision process model by adopting a depth deterministic strategy gradient algorithm improved by introducing a multi-strategy mechanism. Therefore, the robustness of system operation can be greatly improved, multi-energy-flow scheduling is carried out based on the electricity-hydrogen-heat low-carbon building energy system, the cost loss of the battery is effectively reduced, and the operation cost of the system is reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of multi-energy flow scheduling, and particularly relates to a multi-energy flow scheduling method of an electricity-hydrogen-heat low-carbon building energy system and a multi-energy flow scheduling system of the electricity-hydrogen-heat low-carbon building energy system. BACKGROUND

[0002] In the field of integrated energy systems, the integration of renewable energy and clean hydrogen fuel cells is becoming a mainstream trend. However, the inherent source-load uncertainty and the close coupling characteristics between multi-energy flows of this system bring significant technical challenges to the optimization scheduling strategy of the system.

[0003] In the related art, energy storage devices such as battery energy storage systems and heat storage tanks are usually integrated to effectively absorb and utilize renewable energy, thereby improving the scheduling flexibility of the system and achieving a certain degree of fuel consumption reduction, system cost reduction and economic enhancement. However, most of them focus on the analysis of the optimization performance of the system in specific application scenarios. However, in actual operation, the load borne by the system is often difficult to completely match the preset scenario, thus the running robustness of the system is low, and the cost loss of the battery is high during long-term operation, greatly increasing the operation cost of the system. SUMMARY

[0004] To solve the above technical problems, the present application provides a multi-energy flow scheduling method of an electricity-hydrogen-heat low-carbon building energy system, which can greatly improve the robustness of system operation, and effectively reduce the cost loss of the battery based on the multi-energy flow scheduling of the electricity-hydrogen-heat low-carbon building energy system, thereby reducing the operation cost of the system.

[0005] The technical scheme adopted by the present application is as follows:

[0006] A multi-energy flow scheduling method of an electricity-hydrogen-heat low-carbon building energy system, the scheduling method comprising the following steps: S1, establishing physical characteristic models of each related device in the electricity-hydrogen-heat low-carbon building energy system and operation constraint conditions; S2, establishing an objective function with the goal of balancing system operation cost and sustainable operation cost; S3, constructing a Markov decision process model based on the objective function and in combination with the physical characteristic models of each related device and the operation constraint conditions; S4, using a deep deterministic policy gradient algorithm improved by introducing a multi-strategy mechanism to optimize and schedule the multi-energy flow of the electricity-hydrogen-heat low-carbon building energy system based on the Markov decision process model.

[0007] In an embodiment of the present application, the electricity-hydrogen-heat low-carbon building energy system comprises a photovoltaic power generation device, an electrolytic tank, a hydrogen storage tank, a fuel cell, a heat pump and a heat storage tank.

[0008] In one embodiment of the present application, the physical property model of the photovoltaic power generation device is:

[0009] P el = η inv G PV N PV A PV τ

[0010] wherein P el is the output power of the photovoltaic power generation device, τ represents the optical efficiency, A PV represents the total surface area of each photovoltaic component in the photovoltaic power generation device, G represents the solar radiation intensity, N PV is the number of photovoltaic components in the photovoltaic power generation device, η inv represents the photoelectric conversion efficiency, η PV represents the photovoltaic efficiency;

[0011] The physical property model of the electrolytic tank as the accommodation of photovoltaic power generation during the day is:

[0012]

[0013] wherein V e represents the electrolytic voltage, V r represents the reversible potential, i e represents the current density, r1 represents the first temperature ohmic impedance factor, r2 represents the second temperature ohmic impedance factor, d1 represents the first pressure ohmic impedance factor, d2 represents the second pressure ohmic impedance factor, k represents the potential factor, t1 represents the first temperature activation impedance factor, t2 represents the second temperature activation impedance factor, t3 represents the third temperature activation impedance factor, T e and p e respectively represent the electrolytic temperature and the electrolytic pressure, represents the hydrogen production rate, N e represents the number of electrolytic tanks, A e represents the electrolytic area, F and η F respectively represent the Faraday constant and the Faraday efficiency, f 11 represents the first Faraday efficiency factor, f 12 represents the second Faraday efficiency factor, f 21 represents the third Faraday efficiency factor, f 22 represents the fourth Faraday efficiency factor, W e represents the electrolytic power;

[0014] The physical property model of the hydrogen storage tank for storing hydrogen produced from photovoltaic electrolysis during the day and supplying the fuel cell power generation during the night is:

[0015]

[0016] wherein TP t H denotes the hydrogen tank pressure at time t, denotes the hydrogen flow rate at time t, z t denotes the hydrogen compressibility factor at time t, T H and V H denote the hydrogen tank temperature and hydrogen volume, respectively, denotes the hydrogen tank pressure at time t+1;

[0017] The physical property model of the fuel cell is:

[0018]

[0019] wherein, denotes the hydrogen consumption rate of the fuel cell, N fc denotes the number of fuel cells, i fc denotes the current density of the fuel cell, A fc denotes the cell area of the fuel cell, Q fc denotes the heat production of the fuel cell, W fc denotes the power generation of the fuel cell, η H denotes the thermal efficiency of the fuel cell, η E denotes the power generation efficiency of the fuel cell;

[0020] The physical property model of the heat pump is:

[0021]

[0022] wherein COP denotes the energy efficiency ratio of the heat pump, hi denotes the first coefficient, h2 denotes the second coefficient, h3 denotes the third coefficient, ΔT HP denotes the difference between the upgrading temperature and the reference temperature, Q HP denotes the heat that the heat pump can provide, W HP denotes the power consumption of the heat pump;

[0023] The physical property model of the thermal storage tank is:

[0024]

[0025] wherein Q TES,k denotes the heat production of the heat pump at time k, H c denotes the maximum heat storage capacity of the thermal storage tank, TESD k denotes the heat storage state of the thermal storage tank at time k, TESD k+1 denotes the heat storage state of the thermal storage tank at time k+1, and ΔT is the time step.

[0026] In one embodiment of the present application, the operation constraints include: a safe operation constraint and a power balance constraint, wherein,

[0027] The safe operation constraint is:

[0028]

[0029] TESD min ≤TESD t ≤TESD max ,

[0030] TP min ≤TP t ≤TP max ,

[0031] wherein, represents an output power of the fuel cell at time t, and respectively represent a minimum value and a maximum value of the output power of the fuel cell, represents an electricity consumption power of the heat pump at time t, and respectively represent a minimum value and a maximum value of the electricity consumption power of the heat pump, represents an electrolysis power of the electrolyzer at time t, and respectively represent a minimum value and a maximum value of the electrolysis power of the electrolyzer, FC,t represents a hydrogen consumption rate of the fuel cell at time t, and respectively represent a minimum value and a maximum value of the hydrogen consumption rate of the fuel cell, EH,t represents a hydrogen production rate of the electrolyzer at time t, and respectively represent a minimum value and a maximum value of the hydrogen production rate of the electrolyzer, t represents a heat storage state of the heat storage tank at time t, min and TESD max respectively represent a minimum value and a maximum value of the heat storage state of the heat storage tank, t represents a hydrogen storage state of the hydrogen storage tank at time t, min and TP max respectively represent a minimum value and a maximum value of the hydrogen storage state of the hydrogen storage tank;

[0032] The power balance constraint is:

[0033]

[0034] wherein, represents the output power of the fuel cell at time t, represents the photovoltaic power generation power at time t, represents the electrical load of the e-h-h low-carbon building energy system at time t, represents the power consumption of the heat pump at time t, represents the electrolysis power of the electrolyzer at time t, represents the heat production of the fuel cell at time t, represents the heat production of the heat pump at time t, represents the heat release / charge of the heat storage tank at time t, wherein, represents the heat release of the heat storage tank at time t when it is positive, represents the heat charge of the heat storage tank at time t when it is negative, represents the thermal load of the e-h-h low-carbon building energy system at time t, represents the hydrogen production flow rate of the electrolyzer at time t, represents the hydrogen release / charge of the hydrogen storage tank at time t, wherein, represents the hydrogen release flow rate of the hydrogen storage tank at time t when it is positive, represents the hydrogen charge flow rate of the hydrogen storage tank at time t when it is negative, represents the hydrogen consumption flow rate of the fuel cell at time t.

[0035] In an embodiment of the present application, the objective function is:

[0036]

[0037] wherein J represents the objective function, λ1 represents the first adjustable parameter, λ2 represents the second adjustable parameter, λ3 represents the third adjustable parameter, C OM,t represents the system operation and maintenance cost, C T represents the sustainable scheduling cost, k represents the kth device in the e-h-h low-carbon building energy system, OM k represents the unit operation cost of the kth device in the e-h-h low-carbon building energy system, P k,t represents the output power of the kth device at time t, TESD0 represents the heat storage state of the heat storage tank at the beginning of the scheduling period, TESD T represents the heat storage state of the heat storage tank at the end of the scheduling period, TP0 represents the heat storage state of the hydrogen storage tank at the beginning of the scheduling period, TP T represents the heat storage state of the hydrogen storage tank at the end of the scheduling period, T represents the scheduling period.

[0038] In an embodiment of the present application, step S3 specifically comprises:

[0039] The state space is established by the following formula:

[0040]

[0041] Wherein, s t represents the state space, and t represents time;

[0042] The action space is established by the following formula:

[0043]

[0044] Wherein, alpha t represents the action space;

[0045] The reward function is established by the following formula:

[0046]

[0047] Wherein, R represents the reward function, lambda4 represents the fourth adjustable parameter, mu1 represents the fifth adjustable parameter, mu2 represents the sixth adjustable parameter, mu3 represents the seventh adjustable parameter, C out,t represents an additional penalty term, represents a thermal imbalance penalty, represents the hydrogen tank pressure constraint penalty, represents an electrical imbalance penalty.

[0048] In an embodiment of the present application, step S4 specifically comprises: training the improved deep deterministic policy gradient algorithm under uncertainty based on the Markov decision process model to obtain a corresponding policy network model; and optimizing and scheduling the multi-energy flow of the electricity-hydrogen-heat low-carbon building energy system according to the policy network model.

[0049] A multi-energy flow scheduling system of an electricity-hydrogen-heat low-carbon building energy system, comprising: a first establishing unit, configured to establish physical characteristic models of various related devices in the electricity-hydrogen-heat low-carbon building energy system and operating constraint conditions; a second establishing unit, configured to establish an objective function with the goal of balancing system operating cost and sustainable operation cost; a third establishing unit, configured to construct a Markov decision process model based on the objective function and in combination with the physical characteristic models and the operating constraint conditions of the various related devices; and an optimization and scheduling unit, configured to optimize and schedule the multi-energy flow of the electricity-hydrogen-heat low-carbon building energy system based on the Markov decision process model by using a deep deterministic policy gradient algorithm improved by introducing a multi-policy mechanism.

[0050] The present application has the following advantages:

[0051] The application can greatly improve the robustness of system operation, and based on the multi-energy flow scheduling of the electric-hydrogen-thermal low-carbon building energy system, the cost loss of the battery is effectively reduced, thereby reducing the operation cost of the system. BRIEF DESCRIPTION OF DRAWINGS

[0052] Figure 1 A flowchart of the multi-energy flow scheduling method of the electric-hydrogen-thermal low-carbon building energy system of the embodiment of the application.

[0053] Figure 2 A structural schematic diagram of the electric-hydrogen-thermal low-carbon building energy system of one embodiment of the application.

[0054] Figure 3 A source load change graph of the electric-hydrogen-thermal low-carbon building energy system of one embodiment of the application in a typical day scenario.

[0055] Figure 4 A convergence curve graph of the multi-energy flow of the electric-hydrogen-thermal low-carbon building energy system of one embodiment of the application when scheduled by different algorithms.

[0056] Figure 5 A state change graph of the energy storage device (hydrogen storage tank and heat storage tank) of the electric-hydrogen-thermal low-carbon building energy system of one embodiment of the application when scheduled by different algorithms in a typical day scenario.

[0057] Figure 6 A block schematic diagram of the multi-energy flow scheduling system of the electric-hydrogen-thermal low-carbon building energy system of the embodiment of the application. DETAILED DESCRIPTION

[0058] The technical solutions in the embodiments of the application will be clearly and completely described below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, rather than all the embodiments of the application. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative work fall within the protection scope of the application.

[0059] Figure 1 A flowchart of the multi-energy flow scheduling method of the electric-hydrogen-thermal low-carbon building energy system of the embodiment of the application.

[0060] As shown in Figure 1 , the multi-energy flow scheduling method of the electric-hydrogen-thermal low-carbon building energy system of the embodiment of the application can include the following steps:

[0061] S1, a physical characteristic model of each related device in the electric-hydrogen-thermal low-carbon building energy system and an operation constraint condition are established.

[0062] In one embodiment of the application, as shown inFigure 2 As shown, the electricity-hydrogen-heat low-carbon building energy system comprises: a photovoltaic power generation device, an electrolyzer, a hydrogen storage tank, a fuel cell, a heat pump and a heat storage tank, and in addition, the system can further comprise an electric load and a heat load.

[0063] In an embodiment of the present application, the physical property model of the photovoltaic power generation device is:

[0064] P el = η inv G, (1) PV τA PV N PV G, (1)

[0065] wherein P el is the output power of the photovoltaic power generation device; τ represents the optical efficiency, A PV represents the total surface area of each photovoltaic component in the photovoltaic power generation device, G represents the solar radiation intensity, N PV is the number of photovoltaic components in the photovoltaic power generation device, η inv represents the photoelectric conversion efficiency, and η PV represents the photovoltaic efficiency.

[0066] wherein the photovoltaic efficiency η PV can be calculated by the following formula:

[0067] η PV = η ref [1- β ref (T PV -T ref )], (2)

[0068] wherein η ref represents the reference efficiency, β ref represents the temperature reference factor, T PV represents the effective photovoltaic temperature, and T T ref represents the reference temperature.

[0069] The physical property model of the electrolyzer as the consumption of photovoltaic power generation during the day is:

[0070]

[0071] wherein V e represents the electrolysis voltage, V r represents the reversible potential, and i e represents the current.represents current density, r1 represents a first temperature ohmic impedance factor, r2 represents a second temperature ohmic impedance factor, d1 represents a first pressure ohmic impedance factor, d2 represents a second pressure ohmic impedance factor, k represents a potential factor, t1 represents a first temperature activation impedance factor, t2 represents a second temperature activation impedance factor, t3 represents a third temperature activation impedance factor, T e and p e respectively represent electrolysis temperature and electrolysis pressure, represents hydrogen production rate, N e represents electrolytic cell number, A e represents electrolysis area, F and η F respectively represent Faraday constant and Faraday efficiency, f 11 represents a first Faraday efficiency factor, f 12 represents a second Faraday efficiency factor, f 21 represents a third Faraday efficiency factor, f 22 represents a fourth Faraday efficiency factor, W e represents electrolysis power;

[0072] The hydrogen storage tank is used for storing hydrogen produced by photovoltaic electrolysis during the day and supplying fuel cells to generate electricity at night. The physical property model is:

[0073]

[0074] wherein TP t H represents the hydrogen storage tank pressure at t, represents the hydrogen flow at t, z t represents the hydrogen compression coefficient corresponding to t, T H and V H respectively represent the hydrogen storage tank temperature and hydrogen volume, represents the hydrogen storage tank pressure at t+1.

[0075] wherein the hydrogen compression coefficient z t corresponding to t can be calculated by the following formula:

[0076]

[0077] wherein z t (TP t H ) represents the hydrogen compression coefficient corresponding to the hydrogen storage tank pressure at t, a i , b i , c i respectively represent coefficients, and the specific values can be calibrated according to the actual situation.

[0078] The physical property model of the fuel cell is:

[0079]

[0080] wherein, represents the hydrogen consumption rate of the fuel cell, N fc represents the number of fuel cells, i fc represents the current density of the fuel cell, A fc represents the cell area of the fuel cell, Q fc represents the heat production of the fuel cell, W fc represents the power generation of the fuel cell, η H represents the thermal efficiency of the fuel cell, η E represents the power generation efficiency of the fuel cell;

[0081] The physical property model of the heat pump is:

[0082]

[0083] wherein, COP represents the energy efficiency ratio of the heat pump, h1 represents the first coefficient, h2 represents the second coefficient, h3 represents the third coefficient, ΔT HP represents the difference between the upgrading temperature and the reference temperature, Q HP represents the heat that can be provided by the heat pump, W HP represents the power consumption of the heat pump;

[0084] The physical property model of the heat storage tank is:

[0085]

[0086] wherein, Q TES,k represents the heat production of the heat pump at k time, H c represents the maximum heat storage of the heat storage tank, TESD k represents the heat storage state of the heat storage tank at k time, TESD k+1 represents the heat storage state of the heat storage tank at k+1 time, ΔT is a time step.

[0087] S2, a target function is established to balance the system operation cost and the sustainable operation cost.

[0088] In an embodiment of the present application, the operation constraint conditions include: a safe operation constraint and a power balance constraint, wherein,

[0089] The safe operation constraint is:

[0090]

[0091] TESD min ≤ TESD t ≤ TESDmax , (14)

[0092] TP min ≤TP t ≤TP max , (15)

[0093] where, denotes the output power of the fuel cell at time t, and denote the minimum and maximum values of the output power of the fuel cell, respectively, denotes the electricity consumption power of the heat pump at time t, and denote the minimum and maximum values of the electricity consumption power of the heat pump, respectively, denotes the electrolysis power of the electrolyzer at time t, and denote the minimum and maximum values of the electrolysis power of the electrolyzer, respectively, FC,t denotes the hydrogen consumption rate of the fuel cell at time t, and denote the minimum and maximum values of the hydrogen consumption rate of the fuel cell, respectively, EH,t denotes the hydrogen production rate of the electrolyzer at time t, and denote the minimum and maximum values of the hydrogen production rate of the electrolyzer, respectively, t denotes the thermal storage state of the thermal storage tank at time t, min and TESD max denote the minimum and maximum values of the thermal storage state of the thermal storage tank, respectively, t denotes the hydrogen storage state of the hydrogen storage tank at time t, min and TP max denote the minimum and maximum values of the hydrogen storage state of the hydrogen storage tank, respectively;

[0094] The power balance constraints are:

[0095]

[0096] where, denotes the output power of the fuel cell at time t, denotes the photovoltaic power generation at time t, denotes the electrical load of the e-hydrogen-thermal low-carbon building energy system at time t, denotes the electricity consumption power of the heat pump at time t, denotes the electrolysis power of the electrolyzer at time t, denotes the heat production of the fuel cell at time t, denotes the heat production of the heat pump at time t, denotes the heat discharge / charge of the thermal storage tank at time t, is positive, indicating the heat release amount of the heat storage tank at time t, is negative, indicating the heat charging amount of the heat storage tank at time t, represents the heat load of the electric-hydrogen-heat low-carbon building energy system at time t, represents the hydrogen production flow rate of the electrolyzer at time t, represents the hydrogen release amount of the hydrogen storage tank at time t or the hydrogen charging flow rate into the hydrogen storage tank, wherein, is positive, indicating the hydrogen release flow rate of the hydrogen storage tank at time t, is negative, indicating the hydrogen charging flow rate into the hydrogen storage tank at time t, represents the hydrogen consumption flow rate of the fuel cell at time t.

[0097] In an embodiment of the present application, the objective function is:

[0098]

[0099] wherein J represents the objective function, λ1 represents the first adjustable parameter (operation and maintenance cost parameter), λ2 represents the second adjustable parameter (heat storage tank state cost parameter), λ3 represents the third adjustable parameter (hydrogen storage tank state cost parameter), C OM,t represents the system operation and maintenance cost, C T represents the sustainable scheduling cost, k represents the kth device in the electric-hydrogen-heat low-carbon building energy system, OM k represents the unit operation cost of the kth device in the electric-hydrogen-heat low-carbon building energy system, P k,t represents the output power of the kth device at time t, TESD0 represents the heat storage state of the heat storage tank at the beginning of the scheduling period, TESD T represents the heat storage state of the heat storage tank at the end of the scheduling period, TP0 represents the heat storage state of the hydrogen storage tank at the beginning of the scheduling period, TP T represents the heat storage state of the hydrogen storage tank at the end of the scheduling period, T represents the scheduling period.

[0100] S3, based on the objective function and in combination with the physical characteristic model of each related device and the operation constraint condition, a Markov decision process model is constructed.

[0101] In an embodiment of the present application, step S3 specifically comprises:

[0102] The state space is established by the following formula:

[0103]

[0104] wherein s t represents the state space, t represents time;

[0105] The action space is established by the following formula:

[0106]

[0107] wherein, alpha t represents an action space;

[0108] The reward function is established by the following formula:

[0109]

[0110] wherein, R represents a reward function, lambda4 represents a fourth adjustable parameter (an additional penalty parameter), mu1 represents a fifth adjustable parameter (a thermal energy imbalance penalty parameter), mu2 represents a sixth adjustable parameter (a pressure imbalance penalty parameter), mu3 represents a seventh adjustable parameter (an electric imbalance penalty parameter), C out,t represents an additional increased penalty term, represents a thermal imbalance penalty, represents a hydrogen storage tank pressure constraint penalty, represents an electric imbalance penalty.

[0111] In step S4, the improved deep deterministic policy gradient algorithm based on the multi-policy mechanism is used to optimize and dispatch the multi-energy flow of the electric-hydrogen-thermal low-carbon building energy system based on a Markov decision process model.

[0112] In an embodiment of the present application, step S4 specifically comprises: training the improved deep deterministic policy gradient algorithm under uncertainty based on a Markov decision process model to obtain a corresponding policy network model; and optimizing and dispatching the multi-energy flow of the electric-hydrogen-thermal low-carbon building energy system according to the policy network model.

[0113] Specifically, the improved deep deterministic policy gradient algorithm specifically includes: in the framework of the traditional deep deterministic policy gradient algorithm, the OU (Ornstein-Uhlenbeck) noise is introduced to enhance the exploration ability, so as to improve the exploration efficiency of the algorithm in the state space. Then, when the action is executed, the OU noise is introduced at the same time, and the original action (i.e. the action without adding noise) and an adjusted action obtained by directly subtracting the corresponding noise value from the original action are reserved. In order to maintain the accumulated experience information of the original algorithm, the experience obtained by interacting with the environment by the action with noise is set as a must-keep item. Then, the other two actions (the original action and the adjusted action) are also interacted with the environment, and the action experience with the highest reward value is selected from them. Next, the action experience is compared with the experience generated by the action with noise in terms of reward value: if the reward value is higher than the reward value corresponding to the action with noise, the action experience is stored in the experience buffer together with the original experience; otherwise, only the original experience is stored in the experience buffer. Thus, the improved deep deterministic policy gradient algorithm reduces the high-quality action experience that is accidentally ignored due to the addition of noise, while ensuring that the algorithm can capture and utilize these potential beneficial experiences more quickly and effectively, thereby promoting the algorithm to achieve a more efficient and robust balance point between exploration and utilization.

[0114] In an embodiment of the present application, the improved deep deterministic policy gradient algorithm is trained under uncertainty based on a Markov decision process model, which includes: constructing an initial policy network P and an evaluation network Q, and setting up the corresponding target copies: target policy network P' and target evaluation network Q'. The purpose of the policy network is to receive the feature vector input reflecting the system operating conditions, and convert it into executable scheduling decision variables (actions) through an internal mapping mechanism. In contrast, the evaluation network is designed to process the joint input combination of state and action, and its output is the corresponding reward prediction.

[0115] In order to simulate the uncertainty of the source side and the load side encountered by the HES (Hybrid Energy Systems) in the actual operation, a series of source-load scenarios can be randomly generated according to the preset typical daily scene before each training iteration. Specifically, before the start of each training iteration, a typical scene is taken as a benchmark, and no more than 30% of the source-load fluctuation is introduced into it, so as to simulate the uncertainty factors existing in the actual operation environment of the HES. The optimization scheduling strategy is executed under this uncertainty framework, aiming to enhance the robust performance of the algorithm when facing uncertain situations. And at the time node of each training round, a strategy mechanism is implemented to promote environmental exploration and the accumulation of effective training experience. Specifically, when the action is executed, the original action (i.e. the action without adding noise) and an adjusted action obtained by directly subtracting the corresponding noise value from the original action are retained. In order to maintain the experience information accumulated by the original algorithm, the experience obtained by the noisy action interacting with the environment is set as a must-retain item. Then, the other two actions (the original action and the adjusted action) are also interacted with the environment, and the action experience with the highest reward value is selected from them. Next, the reward value of this set of action experience is compared with that of the experience generated by the noisy action: if the reward value is higher than that of the noisy action, the set of action experience and the original experience are stored in the experience buffer together, wherein each set of experience is recorded in the form of a four-tuple, i.e. (s t ,α t ,r t ,s t+1 ), wherein s t represents the state space at time t (i.e. the state at time t), α t represents the action space at time t (i.e. the action at time t), r t represents the reward obtained at time t, and s t+1 represents the state space at time t+1 (i.e. the state at time t+1); otherwise, only the original experience is stored in the experience buffer.

[0116] After completing the training process of the neural network, the trained policy network model is saved for application in the efficient intelligent scheduling task of the photovoltaic power generation-based electric-hydrogen-thermal low-carbon building energy system. In the actual operation scenario, the policy network model can quickly respond to the current state of the system and output an optimized scheduling scheme, aiming to improve energy utilization efficiency and maximize system operation and maintenance costs.

[0117] In a specific embodiment of the present application, the specification parameters and constraint configurations of each device in the electric-hydrogen-thermal low-carbon building energy system can be as shown in Table 1:

[0118] Table 1

[0119]

[0120]

[0121] Furthermore, the main training parameters of the improved deep deterministic policy gradient (I-DDPG) algorithm are listed in Table 2 in detail, as shown in the following table:

[0122] Table 2

[0123]

[0124]

[0125] Among them, in order to achieve a reasonable balance between the system operation and maintenance costs and the additional penalty costs, the system operation and maintenance cost coefficient was carefully adjusted and set to a smaller value, in order to ensure that the system can prioritize the elimination of additional penalty items as the optimization goal, thereby avoiding the risk of additional penalties caused by excessive pursuit of reducing system operation costs. Before the start of each training cycle, based on the load data under typical daily scenarios (such as Figure 3 As shown in the figure, E, H, PV, and T represent electric load, thermal load, photovoltaic power generation and ambient temperature respectively), random source load fluctuations of no more than 30% of the current load are introduced to construct diverse training scenarios. In order to verify the effectiveness of the algorithm improvement measures, the training scenario load, action noise, and initial weight data of the policy network and value network are retained for each training cycle when executing the traditional DDPG (Deep Deterministic Policy Gradient) algorithm. These data are then used as the training load, exploration noise, and initialization parameters of the policy and value networks of the improved DDPG algorithm (i.e., I-DDPG). The training process of the two algorithms is carried out through Figure 4 The training curve in the figure is intuitively displayed, where the "mean" subscript represents the average value of the reward for every 5 steps of training to reflect the performance changes during the training process, that is, DDPG mean and I-DDPG mean Represents the average value of every 5-step training reward in the traditional deep deterministic policy gradient algorithm and the improved deep deterministic policy gradient algorithm. Figure 4The results show that, under the same source load field, noise characteristics and network initialization parameter settings, both algorithms show large negative reward values in the initial exploration stage, which reflects that the algorithm has not found an effective strategy in the preliminary attempt. With the deepening of the training process, thanks to the design of multiple excellent experience storage mechanisms, the I-DDPG algorithm continuously shows a significant advantage over DDPG in reward value. Specifically, the I-DDPG algorithm starts to converge gradually at about 140 training rounds, although there is still some fluctuation due to noise interference, while the DDPG algorithm starts to converge at about 160 rounds. In terms of convergence speed, the I-DDPG algorithm has improved by 12.5% compared with the original algorithm. To further refine the analysis, Figure 4 The last 20 training rounds of reward values are enlarged in the illustration part of FIG. 1, which clearly shows the difference in convergence values between the two algorithms. The results show that the average reward value of the DDPG algorithm in the last 20 rounds is -3.0498, while the average reward value of the I-DDPG algorithm is improved to -2.3542. In terms of convergence average, the I-DDPG algorithm has improved by 22.8% compared with the DDPG algorithm. This significant improvement not only verifies the rationality of the improvement measures proposed in the present application, but also highlights the advantages of the improved algorithm in improving training efficiency and performance. After training, the policy network model is saved and simulated for optimization scheduling under a typical day scenario. Based on this, the comprehensive scheduling cost of the system within 24 hours (including system operation cost and sustainable operation cost) and the final state of the thermal storage tank and hydrogen storage tank are obtained, as shown in Table 3:

[0126] Table 3 Comparison of scheduling results of different algorithms under a typical day scenario

[0127] Parameter Rule-based DDPG I-DDPG Integrated scheduling cost (yuan) 68.41 53.38 51.88 <![CDATA[TESD T ]]> 0.426 0.553 0.56 TP T ]]> 9.64 9.66 9.69

[0128] Table 3 details the comparative analysis of different algorithms on the scheduling results under typical day scenario conditions. In the evaluation of the comprehensive scheduling cost, compared with the traditional deep deterministic policy gradient (DDPG) algorithm, the improved deep deterministic policy gradient (I-DDPG) algorithm realizes a 2.80% cost reduction. At the same time, compared with the rule-based scheduling algorithm (Rule-based), the I-DDPG algorithm realizes a significant reduction of 24.16% in the comprehensive scheduling cost. In terms of the scheduling final state of the energy storage device, for the scheduling effect of the heat storage tank, the performance of the I-DDPG algorithm is improved by 1.2% compared with the DDPG algorithm. Compared with the rule-based algorithm, the I-DDPG algorithm improves the final state of the heat storage tank by 31.45%. In terms of the scheduling of the hydrogen storage tank, the I-DDPG algorithm also shows better performance compared with the DDPG algorithm, achieving a 3.1% improvement. Compared with the rule-based algorithm, the I-DDPG algorithm improves the final state of the hydrogen storage tank by 5.18%.

[0129] In addition, Figure 5 Intuitively, the change trend of the energy storage state of each energy storage device of the two algorithms under the typical day scenario is shown, which provides a strong basis for further analyzing the algorithm performance.

[0130] In summary, according to the multi-energy flow scheduling method of the electric-hydrogen-thermal low-carbon building energy system of the embodiment of the present application, the physical characteristic model and the operation constraint condition of each related device in the electric-hydrogen-thermal low-carbon building energy system are established, a target function is established with the balance of system operation cost and sustainable operation cost as the target, a Markov decision process model is constructed based on the target function and combined with the physical characteristic model and the operation constraint condition of each related device, and a deep deterministic policy gradient algorithm improved by introducing a multi-strategy mechanism is used to optimize and schedule the multi-energy flow of the electric-hydrogen-thermal low-carbon building energy system based on the Markov decision process model. Therefore, not only the robustness of system operation can be greatly improved, but also the multi-energy flow scheduling based on the electric-hydrogen-thermal low-carbon building energy system effectively reduces the cost loss of the battery, thereby reducing the operation cost of the system.

[0131] Corresponding to the above embodiment, the present application also proposes a multi-energy flow scheduling system of an electric-hydrogen-thermal low-carbon building energy system.

[0132] As Figure 6 shown, the multi-energy flow scheduling system of the electric-hydrogen-thermal low-carbon building energy system of the embodiment of the present application can include a first establishing unit 100, a second establishing unit 200, a third establishing unit 300 and an optimization scheduling unit 400.

[0133] The first establishing unit 100 is configured to establish physical characteristic models of various related devices in the electricity-hydrogen-heat low-carbon building energy system and operation constraint conditions; the second establishing unit 200 is configured to establish a target function with the balance of system operation cost and sustainable operation cost as the target; the third establishing unit 300 is configured to construct a Markov decision process model based on the target function and in combination with the physical characteristic models of the various related devices and the operation constraint conditions; and the optimal scheduling unit 400 is configured to perform optimal scheduling on multi-energy flow of the electricity-hydrogen-heat low-carbon building energy system based on the Markov decision process model by using a deep deterministic policy gradient algorithm improved by introducing a multi-strategy mechanism.

[0134] In an embodiment of the present application, the electricity-hydrogen-heat low-carbon building energy system comprises a photovoltaic power generation device, an electrolyzer, a hydrogen storage tank, a fuel cell, a heat pump and a heat storage tank.

[0135] In an embodiment of the present application, the physical characteristic model of the photovoltaic power generation device is:

[0136] P el =η inv η PV τA PV N PV G,

[0137] wherein P el is the output power of the photovoltaic power generation device, τ represents the optical efficiency, A PV represents the total surface area of the photovoltaic components in the photovoltaic power generation device, G represents the solar radiation intensity, N PV is the number of photovoltaic components in the photovoltaic power generation device, η inv represents the photoelectric conversion efficiency, and η PV represents the photovoltaic efficiency.

[0138] The physical characteristic model of the electrolyzer as the consumption of daytime photovoltaic power generation is:

[0139]

[0140] wherein V e represents the electrolysis voltage, V r represents the reversible potential, i e represents the current density, r1 represents the first temperature ohmic impedance factor, r2 represents the second temperature ohmic impedance factor, d1 represents the first pressure ohmic impedance factor, d2 represents the second pressure ohmic impedance factor, k represents the potential factor, t1 represents the first temperature activation impedance factor, t2 represents the second temperature activation impedance factor, t3 represents the third temperature activation impedance factor, T e and p e respectively represent the electrolysis temperature and the electrolysis pressure, represents the hydrogen production rate, N erepresents the number of electrolytic cells, A e represents the electrolytic area, F and η F respectively represent the Faraday constant and the Faraday efficiency, f 11 represents the first Faraday efficiency factor, f 12 represents the second Faraday efficiency factor, f 21 represents the third Faraday efficiency factor, f 22 represents the fourth Faraday efficiency factor, W e represents the electrolytic power;

[0141] The hydrogen storage tank is used for storing hydrogen produced by photovoltaic electrolysis during the day and supplying fuel cells to generate electricity at night. The physical property model is:

[0142]

[0143] wherein TP t H represents the pressure of the hydrogen storage tank at time t, represents the hydrogen flow at time t, z t represents the corresponding hydrogen compression coefficient at time t, T H and V H respectively represent the temperature of the hydrogen storage tank and the hydrogen volume, represents the pressure of the hydrogen storage tank at time t+1;

[0144] The physical property model of the fuel cell is:

[0145]

[0146] wherein, represents the hydrogen consumption rate of the fuel cell, N fc represents the number of fuel cells, i fc represents the current density of the fuel cell, A fc represents the cell area of the fuel cell, Q fc represents the heat production of the fuel cell, W fc represents the power generation of the fuel cell, η H represents the thermal efficiency of the fuel cell, η E represents the power generation efficiency of the fuel cell;

[0147] The physical property model of the heat pump is:

[0148]

[0149] wherein COP represents the energy efficiency ratio of the heat pump, h1 represents the first coefficient, h2 represents the second coefficient, h3 represents the third coefficient, ΔT HP represents the difference between the upgrading temperature and the reference temperature, Q HPIndicates the amount of heat that the heat pump can provide, W HP Indicates the power consumption of the heat pump;

[0150] The physical characteristic model of the thermal storage tank is:

[0151]

[0152] Among them, Q TES,k represents the heat output of the heat pump at time k, H c Indicates the maximum heat storage capacity of the heat storage tank, TESD k represents the heat storage state of the heat storage tank at time k, TESD k+1 It represents the heat storage state of the heat storage tank at time k+1, and ΔT is the time step.

[0153] In one embodiment of the present invention, the operation constraint conditions include: safe operation constraint and power balance constraint, wherein:

[0154] The safe operation constraints are:

[0155]

[0156] TESD min ≤TESD t ≤TESD max ,

[0157] TP min ≤TP t ≤TP max ,

[0158] in, represents the output power of the fuel cell at time t, and Represent the minimum and maximum output power of the fuel cell, represents the electrical power consumed by the heat pump at time t, and Represent the minimum and maximum values ​​of the heat pump power consumption, represents the electrolytic power of the electrolytic cell at time t, and Respectively represent the minimum and maximum values ​​of the electrolytic power of the electrolytic cell, N FC,t represents the hydrogen consumption rate of the fuel cell at time t, and Represent the minimum and maximum values ​​of the fuel cell hydrogen consumption rate, N EH,t represents the hydrogen production rate of the electrolyzer at time t, and They represent the minimum and maximum hydrogen production rates of the electrolyzer, TESD t Indicates the heat storage state of the thermal storage tank at time t, TESDmin and TESD max respectively represent the minimum and maximum of the heat storage state of the heat storage tank, TP t represents the hydrogen storage state of the hydrogen storage tank at t, TP min and TP max respectively represent the minimum and maximum of the hydrogen storage state of the hydrogen storage tank;

[0159] The power balance constraint is:

[0160]

[0161] wherein, represents the output power of the fuel cell at t, represents the photovoltaic power at t, represents the electrical load of the electro-hydro-thermal low-carbon building energy system at t, represents the power consumption of the heat pump at t, represents the electrolysis power of the electrolyzer at t, represents the heat production of the fuel cell at t, represents the heat production of the heat pump at t, represents the heat release / charge of the heat storage tank at t, wherein, represents the heat release of the heat storage tank at t when it is positive, represents the heat charge of the heat storage tank at t when it is negative, represents the heat load of the electro-hydro-thermal low-carbon building energy system at t, represents the hydrogen production flow rate of the electrolyzer at t, represents the hydrogen release / charge of the hydrogen storage tank at t, wherein, represents the hydrogen release of the hydrogen storage tank at t when it is positive, represents the hydrogen charge of the hydrogen storage tank at t when it is negative, represents the hydrogen consumption flow rate of the fuel cell at t.

[0162] In an embodiment of the present application, the objective function is:

[0163]

[0164] wherein, J represents the objective function, λ1 represents the first adjustable parameter, λ2 represents the second adjustable parameter, λ3 represents the third adjustable parameter, C OM,t represents the system operation and maintenance cost, C T represents the sustainable scheduling cost, k represents the kth device in the electro-hydro-thermal low-carbon building energy system, OM k represents the unit operation cost of the kth device in the electro-hydro-thermal low-carbon building energy system, P k,trepresents the output power of the kth device at time t, TESD0 represents the heat storage state of the thermal storage tank at the beginning of the scheduling period, TESD T represents the heat storage state of the thermal storage tank at the end of the scheduling period, TP0 represents the heat storage state of the hydrogen storage tank at the beginning of the scheduling period, TP T represents the heat storage state of the hydrogen storage tank at the end of the scheduling period, T represents the scheduling period.

[0165] In an embodiment of the present application, the third establishing unit 300 is specifically configured to:

[0166] The state space is established by the following formula:

[0167]

[0168] Wherein, s t represents the state space, and t represents time;

[0169] The action space is established by the following formula:

[0170]

[0171] Wherein, a t represents the action space;

[0172] The reward function is established by the following formula:

[0173]

[0174] Wherein, R represents the reward function, λ4 represents the fourth adjustable parameter, μ1 represents the fifth adjustable parameter, μ2 represents the sixth adjustable parameter, μ3 represents the seventh adjustable parameter, C out,t represents an additional penalty term, represents a thermal imbalance penalty, represents a hydrogen storage tank pressure constraint penalty, represents an electrical imbalance penalty.

[0175] In an embodiment of the present application, the optimization scheduling unit 400 is specifically configured to: train the improved deep deterministic policy gradient algorithm under uncertainty based on the Markov decision process model to obtain a corresponding policy network model; and optimize the multi-energy flow of the electricity-hydrogen-heat low-carbon building energy system according to the policy network model.

[0176] It should be noted that the details of the multi-energy flow scheduling system of the electricity-hydrogen-heat low-carbon building energy system of the present application are not disclosed, please refer to the details disclosed in the above-mentioned multi-energy flow scheduling method of the electricity-hydrogen-heat low-carbon building energy system, which will not be described here in detail.

[0177] The multi-energy flow scheduling system of the electricity-hydrogen-heat low-carbon building energy system according to the embodiment of the present application can greatly improve the robustness of system operation, effectively reduce the cost loss of the battery based on the multi-energy flow scheduling of the electricity-hydrogen-heat low-carbon building energy system, and thus reduce the operation cost of the system.

[0178] In the description of the present application, the terms "first", "second" are only for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined with "first", "second" can explicitly or implicitly include one or more of the features. The meaning of "plurality" is two or more, unless otherwise specifically limited.

[0179] In the present application, unless otherwise specifically defined and limited, the terms "mounting", "connection", "connecting", "fixing" and the like should be understood in a broad sense, for example, can be fixed connection, can also be detachable connection, or integral; can be mechanical connection, can also be electrical connection; can be directly connected, can also be indirectly connected through an intermediate medium, can be the internal communication of two elements or the interaction relationship between two elements. For those skilled in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances.

[0180] In the present application, unless otherwise specifically defined and limited, the first feature "on" or "under" the second feature can be that the first and second features are in direct contact, or the first and second features are in indirect contact through an intermediate medium. Moreover, the first feature "above", "over" and "on" the second feature can be that the first feature is directly above or obliquely above the second feature, or only indicates that the horizontal height of the first feature is higher than that of the second feature. The first feature "below", "under" and "under" the second feature can be that the first feature is directly below or obliquely below the second feature, or only indicates that the horizontal height of the first feature is less than that of the second feature.

[0181] In the description of the specification, the description of the terms "one embodiment", "some embodiments", "an example", "a specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. In the specification, the illustrative description of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any appropriate manner in one or more embodiments or examples. In addition, the person skilled in the art can combine and combine the different embodiments or examples described in the specification and the features of the different embodiments or examples without contradiction.

[0182] In addition, each functional unit in each embodiment of the present application can be integrated in one processing module, or each unit can exist physically, or two or more units can be integrated in one module. The integrated module can be realized in the form of hardware or in the form of a software functional module. The integrated module, if realized in the form of a software functional module and sold or used as an independent product, can also be stored in a computer readable storage medium.

[0183] Although the embodiments of the present application have been shown and described above, it should be understood that the above embodiments are exemplary and should not be construed as limiting the present application, and those skilled in the art can make changes, modifications, replacements and variations to the above embodiments within the scope of the present application.

Claims

1. A multi-energy flow scheduling method for an electricity-hydrogen-heat low-carbon building energy system, characterized in that: The scheduling method comprises the following steps: S1, establishing the physical characteristic model and operation constraint conditions of each relevant equipment in the electricity-hydrogen-heat low-carbon building energy system; S2, establishes an objective function with the goal of balancing system operating costs and sustainable operating costs; S3, constructing a Markov decision process model based on the objective function and in combination with the physical characteristic models of the relevant devices and the operating constraints; S4, using the improved deep deterministic policy gradient algorithm with the introduction of a multi-strategy mechanism to optimize the scheduling of the multi-energy flows of the electricity-hydrogen-heat low-carbon building energy system based on the Markov decision process model.

2. The multi-energy flow scheduling method of the electricity-hydrogen-heat low-carbon building energy system according to claim 1 is characterized in that: The electricity-hydrogen-heat low-carbon building energy system includes: photovoltaic power generation equipment, electrolyzer, hydrogen storage tank, fuel cell, heat pump and heat storage tank.

3. The multi-energy flow scheduling method of the electricity-hydrogen-heat low-carbon building energy system according to claim 2 is characterized in that: The physical characteristic model of the photovoltaic power generation equipment is: P el =the inv or PV theA PV N PV G, Among them, P el is the output power of the photovoltaic power generation equipment; τ represents the optical efficiency, A PV represents the total surface area of ​​each photovoltaic module in the photovoltaic power generation equipment, G represents the solar radiation intensity, N PV The number of photovoltaic modules in the photovoltaic power generation equipment, η inv represents the photoelectric conversion efficiency, η PV represents photovoltaic efficiency; The electrolyzer is used to absorb photovoltaic power generation during the day, and the physical characteristic model is: Among them, V e Represents the electrolysis voltage, V r represents the reversible potential, i e represents the current density, r1 represents the first temperature ohmic impedance factor, r2 represents the second temperature ohmic impedance factor, d1 represents the first pressure ohmic impedance factor, d2 represents the second pressure ohmic impedance factor, k represents the potential factor, t1 represents the first temperature activation impedance factor, t2 represents the second temperature activation impedance factor, t3 represents the third temperature activation impedance factor, T e and p e represent the electrolysis temperature and electrolysis pressure respectively, represents the hydrogen yield, N e Indicates the number of electrolytic cells, A e Indicates the electrolysis area, F and η F They represent the Faraday constant and Faraday efficiency, respectively, 11 represents the first Faraday efficiency factor, f 12 represents the second Faraday efficiency factor, f 21 represents the third Faraday efficiency factor, f 22 represents the fourth Faraday efficiency factor, W e Indicates electrolysis power; The hydrogen storage tank is used to store hydrogen produced from photovoltaic electrolysis during the day and supply it to the fuel cell for power generation at night. The physical characteristic model is: Among them, TP t H Indicates the hydrogen storage tank pressure at time t, represents the hydrogen flow rate at time t, z t represents the hydrogen compression coefficient corresponding to time t, T H and V H represent the hydrogen storage tank temperature and hydrogen volume respectively, Indicates the hydrogen storage tank pressure at time t+1; The physical characteristic model of the fuel cell is: in, represents the hydrogen consumption rate of the fuel cell, N fc represents the number of fuel cells, i fc represents the current density of the fuel cell, A fc represents the cell area of ​​the fuel cell, Q fc represents the heat generation of the fuel cell, W fc represents the power generation of the fuel cell, η H represents the thermal efficiency of the fuel cell, η E represents the power generation efficiency of the fuel cell; The physical characteristic model of the heat pump is: Where COP represents the energy efficiency ratio of the heat pump, h1 represents the first coefficient, h2 represents the second coefficient, h3 represents the third coefficient, ΔT HP Indicates the difference between the quality improvement temperature and the reference temperature, Q HP Indicates the amount of heat that the heat pump can provide, W HP represents the power consumption of the heat pump; The physical characteristic model of the heat storage tank is: Among them, Q TES,k represents the heat output of the heat pump at time k, H c Indicates the maximum heat storage capacity of the heat storage tank, TESD k represents the heat storage state of the heat storage tank at time k, TESD k+1 represents the heat storage state of the heat storage tank at time k+1, and ΔT is the time step.

4. The multi-energy flow scheduling method for the electricity-hydrogen-heat low-carbon building energy system according to claim 3 is characterized in that: The operation constraints include: safe operation constraints and power balance constraints, wherein: The safe operation constraints are: TESD min ≤TESD t ≤TESD max , TP min ≤TP t ≤TP max , in, represents the output power of the fuel cell at time t, and represent the minimum and maximum output power of the fuel cell, respectively, represents the electrical power consumed by the heat pump at time t, and represent the minimum and maximum values ​​of the heat pump power consumption, respectively. represents the electrolysis power of the electrolytic cell at time t, and Respectively represent the minimum and maximum values ​​of the electrolytic power of the electrolytic cell, N FC,t represents the hydrogen consumption rate of the fuel cell at time t, and Respectively represent the minimum and maximum values ​​of the hydrogen consumption rate of the fuel cell, N EH,t represents the hydrogen production rate of the electrolyzer at time t, and Respectively represent the minimum and maximum hydrogen production rates of the electrolyzer, TESD t Indicates the heat storage state of the thermal storage tank at time t, TESD min and TESD max Respectively represent the minimum and maximum values ​​of the heat storage state of the heat storage tank, TP t Indicates the hydrogen storage state of the hydrogen storage tank at time t, TP min and TP max Respectively represent the minimum and maximum values ​​of the hydrogen storage state of the hydrogen storage tank; The power balance constraint is: in, represents the output power of the fuel cell at time t, represents the photovoltaic power generation power at time t, represents the electric load of the electricity-hydrogen-heat low-carbon building energy system at time t, represents the electrical power consumed by the heat pump at time t, represents the electrolysis power of the electrolytic cell at time t, represents the heat generation of the fuel cell at time t, represents the heat output of the heat pump at time t, It represents the heat released / charged by the heat storage tank at time t, where: When it is a positive value, it means the heat storage tank releases heat at time t. When it is a negative value, it means the heat storage tank is charged with heat at time t. represents the heat load of the electricity-hydrogen-heat low-carbon building energy system at time t, represents the hydrogen production flow rate of the electrolyzer at time t, represents the amount of hydrogen released from the hydrogen storage tank or the flow rate of hydrogen filled into the hydrogen storage tank at time t, wherein, A positive value indicates the hydrogen flow released by the hydrogen storage tank at time t, A negative value indicates the flow of hydrogen being filled into the hydrogen storage tank. represents the hydrogen consumption flow of the fuel cell at time t.

5. The multi-energy flow scheduling method of the electricity-hydrogen-heat low-carbon building energy system according to claim 4 is characterized in that: The objective function is: Where J represents the objective function, λ1 represents the first adjustable parameter, λ2 represents the second adjustable parameter, λ3 represents the third adjustable parameter, and C OM,t represents the system operation and maintenance cost, C T represents the sustainable dispatch cost, k represents the kth device in the electricity-hydrogen-heat low-carbon building energy system, OM k represents the unit operating cost of the kth device in the electricity-hydrogen-heat low-carbon building energy system, P k,t represents the output power of the kth device at time t, TESD0 represents the heat storage state of the thermal storage tank at the beginning of the scheduling cycle, and TESD T It indicates the heat storage state of the heat storage tank at the end of the scheduling period, TP0 indicates the heat storage state of the hydrogen storage tank at the beginning of the scheduling period, and TP T It represents the heat storage status of the hydrogen storage tank at the end of the scheduling period, and T represents the scheduling period.

6. The multi-energy flow scheduling method for the electricity-hydrogen-heat low-carbon building energy system according to claim 5 is characterized in that: Step S3 specifically includes: The state space is established by the following formula: Among them, s t represents the state space, t represents the time; The action space is established by the following formula: Among them, α t represents the action space; The reward function is established by the following formula: Where R represents the reward function, λ4 represents the fourth adjustable parameter, μ1 represents the fifth adjustable parameter, μ2 represents the sixth adjustable parameter, μ3 represents the seventh adjustable parameter, and C out,t represents the additional penalty term, represents the thermal imbalance penalty, represents the hydrogen tank pressure constraint penalty, Indicates power imbalance penalty.

7. The multi-energy flow scheduling method for the electricity-hydrogen-heat low-carbon building energy system according to claim 6 is characterized in that: Step S4 specifically includes: Based on the Markov decision process model, the improved deep deterministic policy gradient algorithm is trained under uncertainty conditions to obtain a corresponding policy network model; The multi-energy flows of the electricity-hydrogen-heat low-carbon building energy system are optimized and scheduled according to the strategic network model.

8. A multi-energy flow scheduling system for an electricity-hydrogen-heat low-carbon building energy system, characterized in that: include: a first establishing unit, configured to establish a physical characteristic model and operating constraints of each relevant device in the electricity-hydrogen-heat low-carbon building energy system; a second establishing unit, the second establishing unit being configured to establish an objective function with the goal of balancing system operating costs and sustainable operating costs; a third establishing unit, configured to construct a Markov decision process model based on the objective function and in combination with the physical characteristic models of the relevant devices and the operating constraints; An optimization scheduling unit is used to optimize the scheduling of multiple energy flows of the electricity-hydrogen-heat low-carbon building energy system based on the Markov decision process model by adopting a deep deterministic policy gradient algorithm improved by introducing a multi-strategy mechanism.

Images