Intra-day multi-time scale control method for electricity-hydrogen-heat comprehensive energy system

Abstract

The invention discloses an intra-day multi-time scale control method for an electricity-hydrogen-heat comprehensive energy system, relates to the field of comprehensive energy systems, considers the fluctuation of intra-day renewable energy output on hour-level and minute-level time scales, and performs hour-level and minute-level predictive control on the electricity-hydrogen-heat comprehensive energy system, so as to improve the energy efficiency of the electricity-hydrogen-heat comprehensive energy system. Therefore, the utilization rate of the electric hydrogen heat comprehensive energy system on renewable energy power fluctuation is improved. A model comprising a power distribution network, a hydrogen production system and a heat supply system is constructed, the intraday multi-time scale control method of the electric hydrogen heat comprehensive energy system is provided, and intraday efficient dispatching of the electric hydrogen heat comprehensive energy system is achieved.

Description

Technical Field

[0001] The present invention relates to the field of integrated energy systems, and in particular to an intra-day multi-time scale control method for an electric, hydrogen and thermal integrated energy system. Background Art

[0002] As an ideal carbon-free energy, hydrogen is an important way to achieve the "dual carbon" strategy. It is expected that by 2050, the proportion of hydrogen generated by electrolysis will increase from less than 1% at present to 70%. Hydrogen production through intermittent new energy sources such as wind and solar energy will help promote the consumption of clean energy and has obvious low-cost and environmental advantages.

[0003] The safe and stable operation of the electrolytic hydrogen production system has high requirements for power quality. Taking the most widely used alkaline electrolytic hydrogen production as an example, unexpected power fluctuations make it difficult to maintain pressure balance on both sides of the anode and cathode of the electrolytic cell, and it is easy for hydrogen and oxygen gases to pass through the porous diaphragm and mix, and even cause explosions. In addition, it takes a certain amount of time for the electrolyzer to increase the operating temperature to the rated operating temperature. The intermittent fluctuations of renewable energy will also make it difficult for the electrolyzer temperature to maintain the rated temperature, resulting in reduced electrolysis efficiency and even frequent start and stop of the electrolyzer, reducing the service life of the electrolysis equipment. Therefore, for renewable energy power generation and hydrogen production systems, how to make full use of the fluctuating renewable energy hydrogen production has become an important problem in improving renewable energy power generation and hydrogen production systems. In response to this problem, many literatures have considered it from the perspective of uncertainty, and by enhancing the accuracy or robustness of day-ahead scheduling, the impact of renewable energy output fluctuations on the hydrogen production system can be reduced. However, the output of renewable energy is highly random and volatile. Day-ahead prediction and scheduling are difficult to effectively deal with the fluctuations in renewable energy output on a short time scale within the day. These unexpected power fluctuations will have an adverse impact on the operation of the electrolytic hydrogen production system. Therefore, the above problems need to be solved urgently. Summary of the invention

[0004] The present invention focuses on the operation and control problems of the electrolytic hydrogen production system on a short-time scale within a day, and aims to achieve economical and efficient operation of the electric, hydrogen and heat integrated energy system. It constructs a coordinated operation model of the park-level electric, hydrogen and heat integrated energy system including the distribution network, electric hydrogen production, and regional heating network. First, an electrolysis system operation model is constructed that takes into account the electrolysis link, the operating status of the electrolyzer, and the thermal dynamic process of the electrolyzer. On this basis, an operation mode of two-way heat exchange between hydrogen production and the regional heating network is proposed, and a mathematical model of the power grid, hydrogen production, and regional heating system with the electrolyzer as the coupling link is established. Then, based on the hourly and minute-level power forecasts of renewable energy output, a two-layer model predictive control for coordinating the dispatch of electricity, hydrogen, and heat networks is proposed.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] The operation of the electric-hydrogen-heat integrated energy system relies on the two-way heat exchange between the hydrogen production system and the heating system. The two-way heat exchange operation mode is based on the following factors: 1) The optimal production temperature of the electrolyzer is about 90°C, which is usually higher than the return water temperature of the district heating network; 2) The supply water temperature of the district heating network is between 80-110°C, which can heat the electrolyte. Therefore, the temperature difference in the electrolyzer and the district heating network can be exchanged through HE for self-heating. By switching the two valves represented by kv1 and kv2, the direction of the heat flow is changed. When kv1 is open and kv2 is closed, heat is transferred from the electrolyzer to the district heating network. When kv1 is closed and kv2 is opened, the heat flow direction is reversed.

[0007] Based on the above analysis, the bidirectional heat exchange operation model is constructed as follows:

[0008]

[0009] in, is the heat exchanged from the electrolyser to the district heating network, and Binary variables representing the switch status of the two valves.

[0010] ② The realization of multi-time scale control of the electric-hydrogen-heat integrated energy system depends on the modeling of the thermal system and the power system;

[0011] The first step is to build a hydrogen production system model as shown below:

[0012] When the electrolyzer is in production, the relationship between hydrogen production power consumption and hydrogen production can be modeled as:

[0013]

[0014] in, and are the electrolytic cell power and temperature at time t, The power consumed by each battery, N c is the number of batteries, k a ,,k b , k c are the approximate coefficients of the electrolytic cell linear model, I EL t is the electrolysis current, A m is the electrolytic cell membrane area, is the hydrogen production, Δt is the time interval, F is the Faraday constant, z is the number of transferred electrons, is the rated electrolysis power of the electrolytic cell.

[0015] The switching between the operating states of the electrolyzer must meet the following constraints:

[0016]

[0017] Among them, t on and t off are the minimum duration of the production state and the shutdown state respectively. Formulas (4-1) and (4-2) represent the minimum time limits of the production state and the shutdown state.

[0018] The electrolytic cell temperature model is modeled as follows

[0019]

[0020] in, is the heat generated in the hydrogen production process, u tn is the thermoneutral voltage of water electrolysis, C EL is the heat capacity of the electrolytic cell, θ HE is the heat transfer coefficient of the heat exchanger, and are the supply and return water temperatures of the heating network nodes connected to the heat exchanger, is the maximum heat transfer power of the heat exchanger.

[0021] In the second step, the heating network is modeled as follows:

[0022]

[0023] T t,b,end =γ loss L b (T t,b,start -T a )+T a (0-61)

[0024]

[0025] H t,i =C w m(T s,i,t -T r,i,t ) (0-63)

[0026]

[0027] m min ≤m t,b ≤m max (0-66)

[0028] Among them, T a is the ambient temperature, m t,b and m t,n denote the mass flow of branch b and node n, A DHN is the incidence matrix, T t,b,start and T t,b,endRespectively represent the temperature of the starting point and the ending point of branch b, γ loss is the temperature loss coefficient per unit distance, L b is the length of branch b, m t,j is the mass flow rate of branch j, T t,i is the temperature of node i, T t,j,end is the temperature at the end of branch j, and are the supply temperature and return water temperature of node i in the heating network and return water network, respectively, H t,i is the heating load of node i, m min and m max is m b The borders of These are the maximum and minimum temperature limits for the heating network and the return network, respectively.

[0029] In the third step, the power system modeling is as follows:

[0030] The constraints of the power system are constructed based on the linear DistFlow model:

[0031]

[0032] V j,t =V i,t -2(r ij P ij,t +x ij Q ij,t ) (0-69)

[0033]

[0034] V min ≤V j,t ≤V max (0-73)

[0035] Among them, P jk,t and Q jk,t are the active and reactive powers flowing through line jk at time t, δ(j) is the set of endpoints of the branch connected to node j, and P r j,t and Q r j,t are the active and reactive power of renewable energy at node j at time t, and Q D t are active and reactive loads respectively, and are the power imported from and exported from the public grid, and is the upper and lower limits of the power of the interconnection line with the large power grid, r ijand x ij is the resistance and reactance of line ij, V j,t is the square value of the voltage at node j, is the maximum active power of the renewable energy connected to node j, Q r j,min and Q r j,max are the minimum and maximum reactive power of renewable energy at node j, V min and V max are the minimum and maximum values ​​of the square of voltage respectively.

[0036] ③Constructing a two-layer model predictive control method: Δt U and Δt L They represent the control range of the upper-layer model predictive control and the lower-layer model predictive control respectively. The prediction range of the lower-layer controller is the current control range of the upper-layer controller Δt U Based on the hourly power forecast of renewable energy, the scheduling plan provided by the upper controller is in the first lower control range Δt L The lower-level controller will then U The remaining time period of the power plant will be scheduled. Considering the safety limit of hydrogen production and the response speed of the equipment, the operation status of the electrolyzer and the CHP unit provided by the upper controller will be implemented. The remaining variables, including the power of the tie line with the grid, the power and heat output of the CHP unit, and the power and temperature of the electrolyzer, will be adjusted according to the real-time power changes of renewable energy.

[0037] The first step is to build the upper-level model predictive control model

[0038] According to the hourly power forecast of renewable energy, the goal of the upper control is to minimize the net operating cost of the electric, hydrogen and thermal integrated energy system. Its objective function is:

[0039]

[0040] Among them, C ULC t is the net cost of the upper MPC, is the operating cost of the combined heat and power unit, C 0t is the electricity transaction cost with the grid, C H2 t is the revenue from hydrogen sales, C om t is the cost of the electrolyzer standby state, λ NG is the price coefficient of natural gas, are the price coefficients for purchasing electricity from the grid and selling electricity to the grid, λ H2 is the price coefficient of hydrogen sales, λSD is the price coefficient of the electrolytic cell in standby state. The control range of the upper controller is set to 1 hour, and the constraints of the upper controller are given by (0-7)-(4-46).

[0041] The second step is to construct the lower-level model predictive control model as follows:

[0042] The lower controller aims to correct the scheduling plan provided by the upper controller according to the real-time power changes of renewable energy. In the lower controller, the operating status of the electrolyzer and the combined heat and power unit follows the scheduling plan of the upper controller. The objective function of the lower controller includes four parts: 1) the net cost of the electric-hydrogen-heat integrated energy system in the prediction time domain; 2) the penalty cost of the power difference of the grid tie line between the upper controller and the lower controller; 3) the penalty cost of the heat supply difference between the upper controller and the lower controller; 4) the penalty cost of overload operation of the electrolyzer.

[0043] The objective function of the lower controller is expressed as:

[0044]

[0045] Among them, C LLC t,τ represents the objective function of the lower-level controller, t L represents the Lth time interval within the current control range t, the subscript τ represents the τth time interval within the current upper control range, ΔC0 t,τ To adjust the cost with the grid power, P 0+t,τ and P 0-t,τ are the power imported from and exported to the grid, ΔC CHP t,τ To adjust the cost of the CHP unit fuel consumption, ΔP CHP t,τ is the adjusted CHP unit power, ΔC H2 t,τ is the adjusted hydrogen sales revenue, μ1 and μ2 are the penalty coefficients for the tie line power change, ΔH t,τ, is the penalty for the difference in heat supply between the upper and lower controllers, μ3 is the corresponding penalty coefficient for the difference in heat supply, ΔC OL t,τ is the cost of electrolyzer overload operation, λ OL is the cost coefficient of overload operation, k OL t,τ Indicates the overload state of the electrolytic cell.

[0046] Within the prediction range of the lower-level controller, the net operating cost of the electric, hydrogen and heat integrated energy system is given by formula (4-47). In order to make the electric, hydrogen and heat integrated energy system follow the interconnection line power curve provided by the upper-level control, the penalty for interconnection line power changes is included in the cost function through a penalty coefficient. The penalty for interconnection line power changes is given by formula (4-48). The adjustment of the operating cost of the cogeneration unit is given by formula (4-49). The adjustment of hydrogen sales revenue is given by formula (4-50). Due to the large inertia of the centralized heating network, the temperature change of the regional heating network is very small during real-time adjustment. Therefore, the operating model of the regional heating network is simplified. To this end, a penalty term for heat supply changes is added to keep the temperature of the regional heating network consistent with the curve provided by the upper-level controller. The penalty term for heat supply is given by formula (4-51). The operating constraints of the lower-level controller are as follows:

[0047]

[0048] In addition, the overload operation capability of the electrolyzer allows the utilization of excess renewable energy. Therefore, in the lower-level controller, the power constraint of the electrolyzer is modeled as:

[0049]

[0050] Among them, k OL The maximum overload rate allowed for the electrolytic cell.

[0051] The third step is to build a two-layer model predictive control

[0052] At the beginning of each upper controller control range, the power forecast of renewable energy and the operating status of the electrolyzer are updated to the upper controller. Then, an hourly scheduling plan is generated by solving the upper optimization problem. The resulting hourly scheduling plan will be updated in the first Δt of the current upper controller control range. L During the remaining time period, the power changes of renewable energy are measured in real time for the lower-level optimization problem. The lower-level controller will correct the hourly scheduling plan until the control range of the current upper-level controller ends.

[0053] The beneficial effects of the present invention are as follows: the present invention takes into account the volatility of renewable energy output within a day on the hourly and minutely time scales, and performs two-level predictive control on the electric, hydrogen and heat integrated energy system at the hourly and minutely levels to improve the utilization rate of renewable energy power fluctuations in the electric, hydrogen and heat integrated energy system; at the same time, a model including a distribution network, a hydrogen production system and a heating system is constructed, and a multi-time scale control method for the electric, hydrogen and heat integrated energy system within a day is proposed, thereby realizing efficient daily scheduling of the electric, hydrogen and heat integrated energy system. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1This is a structural diagram of the electric, hydrogen and heat integrated energy system proposed in an embodiment of the present invention.

[0055] Figure 2 Schematic diagram of the double-layer model predictive control strategy proposed in an embodiment of the present invention.

[0056] Figure 3 Schematic diagram of the double-layer model predictive control process proposed in an embodiment of the present invention.

[0057] Figure 4 Schematic diagram of the topological structure of a 15-node regional heating network in an embodiment of the present invention.

[0058] Figure 5 1 is a power prediction diagram of renewable energy at different time scales in an embodiment of the present invention.

[0059] Figure 6 This is a time-sharing price chart for purchasing and selling electricity in an embodiment of the present invention.

[0060] Figure 7 This is a real-time electrolysis power comparison diagram in an embodiment of the present invention.

[0061] Figure 8 It is a real-time power comparison diagram of the combined heat and power unit in the embodiment of the present invention.

[0062] Fig. 9 This is a comparison diagram of the actual deviation of the power grid tie line in the example of the present invention. DETAILED DESCRIPTION

[0063] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0064] ① The realization of the coordinated operation mode of waste heat recovery from renewable energy electrolysis hydrogen production and heating network depends on the dynamic modeling of equipment in the thermal system and power system. The thermal system mainly includes heat source, thermal pipeline, thermal network and thermal load. The dynamic modeling of the power system includes power electronic converter model and proton exchange membrane electrolyzer model.

[0065] The coupled heat exchange model of the hydrogen production system and the heating system is as follows:

[0066] The bidirectional heat exchange model of hydrogen production and heating system is as follows:

[0067]

[0068] in, is the heat exchanged from the electrolyser to the district heating network, and The binary variables represent the switch status of the two valves. In view of the nonlinearity of (0-1), a linear transformation is performed based on the Big M method:

[0069]

[0070] Among them, t on and t off are the minimum duration of the production state and the shutdown state respectively. Formulas (4-1) and (4-2) represent the minimum time limits of the production state and the shutdown state.

[0071] The electrolysis hydrogen production system is modeled as follows:

[0072] When the electrolyzer is in production, the relationship between hydrogen production power consumption and hydrogen production can be modeled as:

[0073] P t EL =N c P t cell (0-101)

[0074]

[0075] in, and are the electrolytic cell power and temperature at time t, The power consumed by each battery, N c is the number of batteries, k a ,,k b , k c are the approximate coefficients of the electrolytic cell linear model, I EL t is the electrolysis current, A m is the electrolytic cell membrane area, is the hydrogen production, Δt is the time interval, F is the Faraday constant, z is the number of transferred electrons, is the rated electrolysis power of the electrolytic cell.

[0076] The temperature change of the electrolytic cell is mainly affected by the electrolysis power and the heat exchange power. The heat balance model is as follows:

[0077]

[0078]

[0079] in, is the heat generated in the hydrogen production process, u tn is the thermoneutral voltage of water electrolysis, CEL is the heat capacity of the electrolytic cell, θ HE is the heat transfer coefficient of the heat exchanger, and are the supply and return water temperatures of the heating network nodes connected to the heat exchanger, is the maximum heat transfer power of the heat exchanger.

[0080] The heating system is modeled as follows:

[0081] The district heating network model with constant mass flow is given in (4-25)-(4-28), and the constraints on node temperature and pipe mass flow are given in (4-29)-(4-31).

[0082]

[0083] T t,b,end =γ loss L b (T t,b,start -T a )+T a (0-110)

[0084]

[0085] H t,i =C w m(T s,i,t -T r,i,t ) (0-112)

[0086]

[0087] m min ≤m t,b ≤m max (0-115)

[0088] Among them, T a is the ambient temperature, m t,b and m t,n denote the mass flow of branch b and node n, A DHN is the incidence matrix, T t,b,start and T t,b,end Respectively represent the temperature of the starting point and the ending point of branch b, γ loss is the temperature loss coefficient per unit distance, L b is the length of branch b, m t,j is the mass flow rate of branch j, T t,i is the temperature of node i, T t,j,end is the temperature at the end of branch j, and are the supply temperature and return water temperature of node i in the heating network and return water network, respectively, Ht,i is the heating load of node i, m min and m max is m b The borders of These are the maximum and minimum temperature limits for the heating network and the return network, respectively.

[0089] The power system is modeled as follows:

[0090] The constraints of the power system are constructed based on the linear DistFlow model:

[0091]

[0092] V j,t =V i,t -2(r ij P ij,t +x ij Q ij,t ) (0-118)

[0093]

[0094] V min ≤V j,t ≤V max (0-122)

[0095] Among them, P jk,t and Q jk,t are the active and reactive powers flowing through line jk at time t, δ(j) is the set of endpoints of the branch connected to node j, and P r j,t and Q r j,t are the active and reactive power of renewable energy at node j at time t, and Q D t are active and reactive loads respectively, and are the power imported from and exported from the public grid, and is the upper and lower limits of the power of the interconnection line with the large power grid, r ij and x ij is the resistance and reactance of line ij, V j,t is the square value of the voltage at node j, is the maximum active power of the renewable energy connected to node j, Q r j,min and Q r j,max are the minimum and maximum reactive power of renewable energy at node j, V min and V maxare the minimum and maximum values ​​of the square of voltage respectively.

[0096] The operating constraints of the CHP unit are as follows:

[0097]

[0098] in, represents the thermal power, a CHP 、b CHP d CHP is a vector describing the feasible region of the CHP unit power and heat output, is a binary variable used to indicate the start and stop status of the cogeneration unit. is the natural gas consumption rate, η CHP is efficiency, L NG is the lower calorific value of natural gas, ρ CHP for and The proportionality factor between .

[0099] See also Figure 1 The double-layer model predictive control method proposed in the present invention. Δt U and Δt L They represent the control ranges of the upper-layer model predictive control and the lower-layer model predictive control respectively. The prediction range of the lower-layer controller is the current control range of the upper-layer controller Δt U For the remaining time period, based on the hourly power forecast of renewable energy, the scheduling plan provided by the upper controller is within the first lower control range Δt L The lower-level controller will then execute the U The remaining time period will be scheduled. Taking into account the safety limitations of hydrogen production and the response speed of the equipment, the operating status of the electrolyzer and the cogeneration unit provided by the upper controller will be implemented, and the remaining variables, including the power of the interconnection line with the grid, the power and heat output of the cogeneration unit, and the power and temperature of the electrolyzer, will be adjusted according to the real-time power changes of renewable energy.

[0100] According to the hourly power forecast of renewable energy, the goal of the upper control is to minimize the net operating cost of the electric, hydrogen and thermal integrated energy system. Its objective function is:

[0101]

[0102] Among them, C ULC t is the net cost of the upper MPC, C CHP t is the operating cost of the combined heat and power unit, C0 t is the electricity transaction cost with the grid, C H2t is the revenue from hydrogen sales, C om t is the cost of the electrolyzer standby state, λ NG is the price coefficient of natural gas, are the price coefficients for purchasing electricity from the grid and selling electricity to the grid, λ H2 is the price coefficient of hydrogen sales, λ SD is the price coefficient of the electrolytic cell in standby state. The control range of the upper controller is set to 1 hour, and the constraints of the upper controller are given by (0-7)-(4-46).

[0103] The lower controller aims to correct the scheduling plan provided by the upper controller according to the real-time power changes of renewable energy. In the lower controller, the operating status of the electrolyzer and the combined heat and power unit follows the scheduling plan of the upper controller. The objective function of the lower controller includes four parts: 1) the net cost of the electric-hydrogen-heat integrated energy system in the prediction time domain; 2) the penalty cost of the power difference of the grid tie line between the upper controller and the lower controller; 3) the penalty cost of the heat supply difference between the upper controller and the lower controller; 4) the penalty cost of overload operation of the electrolyzer.

[0104] The objective function of the lower controller is expressed as:

[0105]

[0106] Among them, C LLC t,τ represents the objective function of the lower controller, t L represents the Lth time interval within the current control range t, the subscript τ represents the τth time interval within the current upper control range, ΔC0 t,τ To adjust the cost with the grid power, P 0+t,τ and P 0-t,τ are the power imported from and exported to the grid, ΔC CHP t,τ To adjust the cost of the CHP unit fuel consumption, ΔP CHP t,τ is the adjusted CHP unit power, ΔC H2 t,τ is the adjusted hydrogen sales revenue, μ1 and μ2 are the penalty coefficients for the tie line power change, ΔH t,τ, is the penalty for the difference in heat supply between the upper and lower controllers, μ3 is the corresponding penalty coefficient for the difference in heat supply, ΔC OL t,τ is the operating cost of the electrolyzer overload, OL is the cost coefficient of overload operation, k OL t,τIndicates the overload state of the electrolytic cell.

[0107] The operating constraints of the lower-level controller are as follows:

[0108]

[0109]

[0110] In addition, the overload operation capability of the electrolyzer allows the utilization of excess renewable energy. Therefore, in the lower-level controller, the power constraint of the electrolyzer is modeled as:

[0111]

[0112] Among them, k OL The maximum overload rate allowed for the electrolytic cell.

[0113] See also Figure 3 At the beginning of each upper controller control range, the power forecast of renewable energy and the operating status of the electrolyzer will be updated to the upper controller. Then, an hourly scheduling plan is generated by solving the upper optimization problem. The hourly scheduling plan is generated at the first Δt of the current upper controller control range. L During the remaining time period, the power changes of renewable energy are measured in real time for the lower-level optimization problem. The lower-level controller will correct the hourly scheduling plan until the control range of the current upper-level controller ends.

[0114] See also Figure 4 The test was conducted on a 15-node district heating network in . The combined heat and power unit was connected to node 1 of the supply network. The alkaline electrolyzer was configured at node 1 of the district heating network. The operating parameters of the electrolyzer and the electric-hydrogen-heat integrated energy system are shown in Tables 1 and 2. The hourly and 5-minute power forecasts of photovoltaic and wind power are shown in Figure 5 The time period prices for purchasing and selling electricity from the power grid are shown in Figure 6 Since the heating pipes are usually placed underground, it is assumed that the ambient temperature T a is a constant.

[0115] Table 1 Electrolyzer operating parameters

[0116]

[0117] Table 2 Operation parameters of the electric, hydrogen and heat integrated energy system

[0118]

[0119] In order to verify the proposed scheduling method, two scenarios, single-layer model predictive control and double-layer model predictive control, were configured. In scenario 1, the electric-hydrogen-heat integrated energy system is scheduled using hourly single-layer model predictive control, and the power grid acts as a regulator to balance the real-time power fluctuations of renewable energy. Scenario 2 uses double-layer model predictive control for control. Therefore, the effectiveness of double-layer model predictive control is verified by setting scenarios 1 and 2. When the real-time output is greater than the hourly forecast, the excess renewable energy will be cut; when the output is less than the forecast, the power grid makes up for the insufficient renewable energy. In scenario 2, the bidirectional heat exchange and double-layer model predictive control of the hydrogen production and heating systems are considered at the same time.

[0120] Figure 7 The electrolysis power in scenario 1 and scenario 2 is shown. In scenario 1, as the hourly scheduling progresses, the electrolysis power curve changes only at the beginning of the hour. Compared with scenario 1, the electrolysis power in scenario 2 is scheduled every 5 minutes by the two-layer model predictive control, so the electrolysis power can be adjusted in time to follow the real-time power changes of renewable energy. For example, during the period from 0:00 to 1:00, the actual output of renewable energy is much lower than the hourly prediction. In scenario 1, the electrolysis power remains at the hourly scheduling value. In scenario 2, the electrolysis power is reduced by about 1MW to follow the power changes of renewable energy. Therefore, through the two-layer model predictive control, the hydrogen produced by water electrolysis effectively alleviates the real-time power fluctuations of renewable energy.

[0121] Figure 8 The output power of the CHP unit in scenario 1 and scenario 2 is shown. Similar to the electrolysis power, the power of the CHP unit in scenario 1 remains constant at the hourly level. Since the main power fluctuations of renewable energy are regulated by the electrolyzer in scenario 2, the power curve of the CHP unit is similar to that in scenario 1. However, during the period of large fluctuations in renewable energy power, the power of the CHP unit in scenario 2 changes to mitigate the fluctuations in renewable energy power. For example, at 5:00, the power of the CHP unit rises to 0.5MW. In addition, there is a stable difference between scenario 1 and scenario 2 during a specific time period. For example, during the period from 13:00 to 14:00, the power of the CHP unit in scenario 2 is about 0.5MW lower than that in scenario 1. This is because the electrolysis power exceeds the hourly dispatch value during the period from 12:00 to 13:00. Therefore, more heat is transferred from the electrolyzer to the district heating network. Therefore, the CHP unit needs to reduce its output to balance the supply and consumption of heat.

[0122] In order to demonstrate the effect of two-layer model predictive control on power grid disturbance suppression, Fig. 9The actual difference in tie-line power with the grid is shown. Since there is no response to real-time power changes of renewable energy in scenario 1, the grid acts as a regulator when the actual output of renewable energy is lower than the hourly forecast value. Therefore, the tie-line power difference in scenario 1 is non-negative throughout the dispatch period. Compared with scenario 1, the tie-line power difference in scenario 2 is significantly reduced, which shows that the two-layer model predictive control reduces the interference of intermittent renewable energy to the grid. Although there is still a power difference in a specific time period, the maximum value is reduced from 1.9MW to 0.8MW, and most of the slight renewable energy power fluctuations are smoothed. This shows that the electric-hydrogen-heat integrated energy system using two-layer model predictive control helps to smooth the tie-line power and reduce the interference of intermittent renewable energy to the grid.

[0123] The operating costs of scenarios 1 and 2 are shown in Table 4. The net cost includes grid power transaction costs, fuel costs for the CHP units, hydrogen sales revenue, and differential penalties associated with grid tie-line power and district heating network heat supply. Compared with scenario 1, the grid power transaction costs and penalties in scenario 2 are similar to those in scenario 1. This is because the short-term power fluctuations of renewable energy tend to a random distribution with zero deviation. Considering the allowance of short-term overload, the hydrogen sales revenue in scenario 2 is ￥1467 more than that in scenario 1. Through the two-layer model predictive control, the differential penalty for grid tie-line power and district heating network heat supply is reduced from ￥7878 to ￥1158. Therefore, the real-time dispatch achieved by the two-layer model predictive control reduces the net cost by about 14.4%.

[0124] Table 4 Operating costs of the electric, hydrogen and heat integrated energy system (￥)

[0125]

[0126] Among the above simulation methods, what is original to the present invention, has never been disclosed, and its working method is different from any existing literature records: the present invention improves the utilization level of energy by establishing an operation mode of recovering waste heat from hydrogen production by electricity for heating, and taking into account the differences in the dynamic characteristics of the power system and the thermal system involved, a parallel multi-rate simulation algorithm is proposed to achieve efficient and accurate simulation of the dynamic processes existing in the coupled system.

[0127] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

[0128] In addition, it should be understood that although the present specification is described according to implementation modes, not every implementation mode contains only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment may also be appropriately combined to form other implementation modes that can be understood by those skilled in the art.

Claims

1. A multi-time scale control method for an electric, hydrogen and heat integrated energy system within a day, characterized in that: The steps include: ① The bidirectional heat exchange operation model of hydrogen production and heating system is constructed as follows: in, is the heat exchanged from the electrolyser to the district heating network, and Binary variables representing the switch status of the two valves kv1 and kv2 respectively; ② The hydrogen production system model is constructed as follows: When the electrolyzer is in production, the relationship between hydrogen production power consumption and hydrogen production is modeled as: P t EL =N c P t cell (0-3) in, and are the electrolytic cell power and temperature at time t, The power consumed by each battery, N c is the number of batteries, k a ,,k b , k c are the approximate coefficients of the electrolytic cell linear model, I ELt is the electrolysis current, A m is the electrolytic cell membrane area, is the hydrogen production, Δt is the time interval, F is the Faraday constant, z is the number of transferred electrons, is the rated electrolysis power of the electrolytic cell; The switching between the operating states of the electrolyzer satisfies the following constraints: Among them, t on and t off are the minimum duration of the production state and the shutdown state, respectively. Formulas (4-1) and (4-2) represent the minimum time limits of the production state and the shutdown state; The electrolytic cell temperature model is modeled as follows: in, is the heat generated in the hydrogen production process, u tn is the thermoneutral voltage of water electrolysis, C EL is the heat capacity of the electrolytic cell, θ HE is the heat transfer coefficient of the heat exchanger, and are the supply and return water temperatures of the heating network nodes connected to the heat exchanger, is the maximum heat transfer power of the heat exchanger; The heating network is modelled as follows: T t,b,end =γ loss L b (T t,b,start -T a )+T a (0-14) H t,i =C w m(T s,i,t -T r,i,t ) (0-16) m min ≤m t,b ≤m max (0-19) Among them, T a is the ambient temperature, m t,b and m t,n denote the mass flow of branch b and node n, A DHN is the incidence matrix, T t,b,start and T t,b,end Respectively represent the temperature of the starting point and the ending point of branch b, γ loss is the temperature loss coefficient per unit distance, L b is the length of branch b, m t,j is the mass flow rate of branch j, T t,i is the temperature of node i, T t,j,end is the temperature at the end of branch j, and are the supply temperature and return water temperature of node i in the heating network and return water network, respectively, H t,i is the heating load of node i, m min and m max is m b The borders of The maximum and minimum temperature limits for the heating network and return network, respectively; The power system is modeled as follows: The constraints of the power system are constructed based on the linear DistFlow model: V j,t =V i,t -2(r ij P ij,t +x ij Q ij,t ) (0-22) In min ≤V j,t ≤V max (0-26) Among them, P jk,t and Q jk,t are the active and reactive powers flowing through line jk at time t, δ(j) is the set of endpoints of the branch connected to node j, and P rj,t and Q rj,t are the active and reactive power of renewable energy at node j at time t, and are active and reactive loads respectively, and are the power imported from and exported from the public grid, and is the upper and lower limits of the power of the interconnection line with the large power grid, r ij and x ij is the resistance and reactance of line ij, V j,t is the square value of the voltage at node j, is the maximum active power of the renewable energy connected to node j, Q r j,min and Q r j,max are the minimum and maximum reactive power of renewable energy at node j, V min and V max are the minimum and maximum values ​​of the square of voltage respectively; ③Constructing a two-layer model predictive control method: Δt U and Δt L They represent the control ranges of the upper-layer model predictive control and the lower-layer model predictive control, respectively. The prediction range of the lower-layer controller is the current control range of the upper-layer controller Δt U For the remaining time period, based on the hourly power forecast of renewable energy, the scheduling plan provided by the upper controller is within the first lower control range Δt L The lower-level controller will then U The remaining time period of the power plant will be scheduled. Considering the safety limit of hydrogen production and the response speed of the equipment, the operation status of the electrolyzer and the cogeneration unit provided by the upper controller will be implemented. The remaining variables, including the power of the tie line with the grid, the power and heat output of the cogeneration unit, and the power and temperature of the electrolyzer, will be adjusted according to the real-time power changes of renewable energy. The first step is to build the upper-level model predictive control model According to the hourly power forecast of renewable energy, the goal of the upper control is to minimize the net operating cost of the electric, hydrogen and thermal integrated energy system. Its objective function is: Among them, C ULCt is the net cost of the upper MPC, is the operating cost of the combined heat and power unit, C 0t is the electricity transaction cost with the grid, C H2t is the revenue from hydrogen sales, C omt is the cost of the electrolyzer standby state, λ NG is the price coefficient of natural gas, are the price coefficients for purchasing electricity from the grid and selling electricity to the grid, λ H2 is the price coefficient of hydrogen sales, λ SD is the price coefficient of the electrolytic cell standby state, the control range of the upper controller is set to 1 hour, and the constraints of the upper controller are given by (0-7)-(4-46); The second step is to construct the lower-level model predictive control model as follows: The lower-level controller can correct the scheduling plan provided by the upper-level controller according to the real-time power changes of renewable energy. In the lower-level controller, the operating status of the electrolyzer and the combined heat and power unit follows the scheduling plan of the upper-level controller. The objective function of the lower-level controller includes four parts: 1) the net cost of the electric-hydrogen-heat integrated energy system in the predicted time domain; 2) the penalty cost of the power difference of the grid interconnection line between the upper-level controller and the lower-level controller; 3) the penalty cost of the heat supply difference between the upper-level controller and the lower-level controller; 4) the penalty cost of overload operation of the electrolyzer; The objective function of the lower controller is expressed as: Among them, C LLCt,τ represents the objective function of the lower-level controller, t L represents the Lth time interval within the current control range t, the subscript τ represents the τth time interval within the current upper control range, ΔC 0t,τ To adjust the cost with the grid power, P 0+t,τ and P 0-t,τ are the power imported from and exported to the grid, ΔC CHPt,τ To adjust the cost of the CHP unit fuel consumption, ΔP CHPt,τ is the adjusted CHP unit power, ΔC H2t,τ is the adjusted hydrogen sales revenue, μ1 and μ2 are the penalty coefficients for the tie line power change, ΔH t,τ, is the penalty for the difference in heat supply between the upper and lower controllers, μ3 is the corresponding penalty coefficient for the difference in heat supply, ΔC OLt,τ is the operating cost of the electrolyzer overload, λ OL is the overload operation cost coefficient, k OLt,τ Indicates the overload state of the electrolytic cell; Within the prediction range of the lower controller, the net operating cost of the electric-hydrogen-heat integrated energy system is given by formula (4-47). The penalty for the change in the interconnection line power is included in the cost function through the penalty coefficient. The penalty for the change in the interconnection line power is given by formula (4-48). The adjustment of the operating cost of the cogeneration unit is given by formula (4-49). The adjustment of the hydrogen sales revenue is given by formula (4-50). The operation model of the regional heating network is simplified, and the penalty term for the change in heat supply is added to keep the temperature of the regional heating network consistent with the curve provided by the upper controller. The penalty term for heat supply is given by formula (4-51). The operating constraints of the lower controller are as follows: In the lower level controller, the power constraint of the electrolyzer is modeled as: Among them, k OL is the maximum overload rate allowed for the electrolytic cell; The third step is to build a two-layer model predictive control At the beginning of each upper controller control range, the power forecast of renewable energy and the operating status of the electrolyzer will be updated to the upper controller. Then, an hourly scheduling plan is generated by solving the upper optimization problem. The hourly scheduling plan is generated in the first Δt of the current upper controller control range. L The time period is executed, and in the remaining time period, the power changes of renewable energy are measured in real time for the lower-level optimization problem until the control range of the current upper-level controller ends. The lower-level controller will correct the hourly scheduling plan.

Images