Multi-time-scale distributed predictive control method for electrical hydrogen coupling comprehensive energy system

Abstract

The invention relates to a multi-time-scale distributed predictive control method for an electricity-hydrogen coupled comprehensive energy system. The method comprises the following steps: constructing an off-grid electricity-hydrogen coupled combined cooling heating and power system based on hydrogen-doped natural gas; according to the dynamic response characteristic difference, the off-grid electric hydrogen coupling combined cooling heating and power system is divided into an electric side subsystem, a cold side subsystem and a hot side subsystem; a multi-time-scale distributed prediction control mode is adopted, a local prediction controller is designed, corresponding optimization control is carried out on an electric side subsystem, a cold side subsystem and a hot side subsystem, and the working state of the off-grid electric hydrogen coupling combined cooling heating and power system is correspondingly adjusted. Compared with the prior art, the problems of strong coupling and large time scale span of the system can be effectively solved, the control precision and response speed of the electrical hydrogen coupling comprehensive energy system are improved, the response speed of the electrical side can be increased while the control accuracy of the temperature side is ensured, and computing resources are saved.

Description

Technical Field

[0001] The present invention relates to the field of integrated energy system control technology, and in particular to a multi-time scale distributed predictive control method for an electric-hydrogen coupled integrated energy system. Background Art

[0002] In the global energy transition, hydrogen compressed natural gas (HCNG) has attracted much attention as a new type of low-carbon energy carrier that significantly reduces carbon emissions by adding hydrogen. The introduction of hydrogen-blended natural gas in the integrated energy system has enriched the energy supply form, promoted the consumption of renewable energy, and improved the system's energy efficiency and environmental performance. However, the HCNG electric-hydrogen coupled energy system is highly coupled and has a large difference in the response time of cooling, heating and electricity, which brings new challenges to the system operation and control, especially in multi-time scale coordinated control.

[0003] At present, researchers at home and abroad are mainly focusing on the low carbon, economic and safety of hydrogen-blended natural gas to optimize scheduling. Although existing literature has explored the impact of hydrogen blending strategies, coordinated control of integrated energy systems and seasonal changes on the scheduling of electric hydrogen coupled energy systems, there are few studies on the optimization control of trigeneration integrated energy systems with HCNG as energy input. It is difficult to solve the problems of strong system coupling and large time scale span, and it is impossible to ensure the timeliness and accuracy of system control. Summary of the invention

[0004] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and to provide a multi-time scale distributed predictive control method for an electric-hydrogen coupled integrated energy system, which can effectively solve the problems of strong system coupling and large time scale span, and improve the control accuracy and response speed of the electric-hydrogen coupled integrated energy system.

[0005] The purpose of the present invention can be achieved by the following technical solution: A multi-time scale distributed predictive control method for an electric hydrogen coupled integrated energy system, comprising the following steps:

[0006] S1. Build an off-grid electric hydrogen-coupled combined heating and cooling power system based on hydrogen-blended natural gas;

[0007] S2. According to the differences in dynamic response characteristics, the off-grid electric hydrogen coupled combined heat and power system is divided into three subsystems: the power side, the cold side, and the heat side;

[0008] S3. Adopt multi-time scale distributed predictive control method, design local predictive controller, perform corresponding optimization control on the three subsystems of power side, cold side and hot side, and adjust the working state of off-grid electric hydrogen coupled combined heat and power system accordingly.

[0009] Furthermore, the off-grid electric hydrogen coupled combined heat and power system in step S1 includes a micro-turbine, a double-effect lithium bromide absorption refrigerator, a fuel cell, a cold storage tank, a battery, and a hydrogen storage tank.

[0010] Furthermore, the operation process of the off-grid electric hydrogen coupled cooling, heating and power cogeneration system in step S1 is as follows:

[0011] After HCNG passes through the separation device, natural gas and hydrogen are supplied to the micro-turbine and fuel cell respectively. In the micro-turbine, natural gas is directly sent to the combustion chamber, which produces high-temperature and high-pressure flue gas, driving the gas turbine and generator to convert chemical energy into electrical energy. At the same time, the waste heat flue gas is used for heating and cooling in the double-effect lithium bromide absorption refrigeration machine;

[0012] In the fuel cell, hydrogen reacts with oxygen in the air to generate electricity, and the high-temperature gas discharged exchanges heat with the heating return water to generate hot water;

[0013] Batteries are used as energy storage units for energy balance in dynamic processes;

[0014] As an energy storage unit, the cold storage tank is responsible for regulating the cooling flow rate to ensure stable cooling supply;

[0015] The hydrogen storage tank changes the hydrogen flow rate at the fuel cell inlet by adjusting the charge and discharge.

[0016] Furthermore, in step S2, the control variables of the power-side subsystem are the volume flow rate of the hydrogen-mixed natural gas and the hydrogen flow rate at the fuel cell inlet, and the controlled variable is the user-side power;

[0017] The control variables of the cold side subsystem are the opening of the refrigerant valve of the refrigerator and the cold release flow of the cold storage tank, and the controlled variable is the cooling temperature;

[0018] The control variables of the heat side subsystem are the opening of the micro-turbine regenerative valve and the fuel cell inlet air flow rate; the controlled variable is the heating temperature.

[0019] Furthermore, the state space expression of the electric side subsystem is:

[0020]

[0021] Among them, T N is the discrete sampling time of the electrical subsystem #N, V mix is the flow rate of hydrogen-mixed natural gas, m H2 is the hydrogen flow rate at the fuel cell inlet, and k is the sampling step length.

[0022] Furthermore, the state space expression of the cold side subsystem is:

[0023]

[0024] Among them, T T is the discrete sampling time of the cold side subsystem #C, μ hgr is the opening of the refrigerant valve of the refrigerator, m cold,CS It is the cold release flow of the cold storage tank.

[0025] Furthermore, the state space expression of the hot side subsystem is:

[0026]

[0027] Among them, T T is the discrete sampling time of the hot side subsystem #H, μ reg is the opening of the micro-turbine regenerative valve, m air is the fuel cell inlet air flow rate.

[0028] Furthermore, step S3 includes the following process:

[0029] S31, for the three subsystems of the power side, the cold side and the hot side, respectively design a corresponding power side controller, a cold side controller and a hot side controller, wherein the power side controller, the cold side controller and the hot side controller respectively adopt a first local cost function, a second local cost function and a third local cost function;

[0030] S32, recording the time for simultaneously updating the cold side controller, the hot side controller and the electric side controller as a synchronous sampling period;

[0031] The time when only the power-side controller is updated is recorded as the asynchronous sampling period;

[0032] S33, adopting a multi-time scale distributed information interaction mechanism, in an asynchronous sampling period, the power-side controller takes minimization of the first local cost function as the control target and updates the control quantity of the power-side subsystem accordingly;

[0033] In the synchronous sampling period, the electric side controller, the cold side controller and the hot side controller work together, taking the minimization of the corresponding local cost function as the control objective, and updating the control quantities of the three subsystems of the electric side, the cold side and the hot side accordingly.

[0034] Furthermore, the first local cost function is specifically:

[0035]

[0036] st

[0037]

[0038] Y 1,p (k+i)=Y 1,m (k+i)+h 1 (Y 1(k)-Y 1,m (k))

[0039]

[0040] Among them, N p is the prediction time domain of the electric side subsystem, N c is the control time domain of the electric side subsystem, Q 1,i is the error weighting coefficient of the electric side subsystem, R 1,j is the control weighting coefficient of the electric side subsystem, ΔU 1 =[ΔV mix ,Δm H2 ] T is the control action increment of the electric side subsystem, ΔU 1,max , ΔU 1,min They represent the upper and lower limits of the control action increment of the electric side subsystem, U 1,max , U 1,min They represent the upper and lower limits of the control action of the electric side subsystem, Y 1,p The system prediction output for the electric side subsystem; Y 1,p =N userp is the predicted output of the electric side subsystem after feedback correction, Y 1,r =N userr is the expected trajectory of the electric side subsystem, h is the correction coefficient, Y 1,m (k+i) is the multi-step prediction output of the electric side subsystem, R 1 is the set value, ψ is the softening factor, 0<ψ<1, which is used to make the output Y of the power side subsystem 1 Smooth transition to set value;

[0041] The second local cost function is specifically:

[0042]

[0043] st

[0044]

[0045] Among them, Y 2,p is the system predicted output of the cold side subsystem, Y 2,r is the expected trajectory of the cold side subsystem, ΔU 2 =[Δμ hgr ,Δm cold,CS ] is the control action increment of the cold side subsystem, Q 2,i is the error weighting coefficient of the cold side subsystem, R 2,j is the control weighting coefficient of the cold side subsystem, ΔU 2,max , ΔU 2,min They represent the upper and lower limits of the control action increment of the cold side subsystem, U2,max , U 2,min They represent the upper and lower limits of the control action of the cold-side subsystem respectively;

[0046] The third local cost function is specifically:

[0047]

[0048] st

[0049]

[0050] Among them, Y 3,p is the system predicted output of the hot side subsystem, Y 3,r is the expected trajectory of the hot side subsystem, ΔU 3 =[Δμ reg ,Δm air ] is the control action increment of the hot side subsystem, Q 3,i is the error weighting coefficient of the hot side subsystem, R 3,j is the control weighting coefficient of the hot side subsystem, ΔU 3,max , ΔU 3,min They represent the upper and lower limits of the control action increment of the hot side subsystem, U 3,max , U 3,min They represent the upper and lower limits of the control action of the hot side subsystem respectively.

[0051] Furthermore, the specific process of step S33 is as follows:

[0052] In the asynchronous sampling period k a The electric side controller uses the cold side and hot side controllers in the previous temperature side sampling cycle T T The calculated control quantity U 2 and U 3 , correspondingly calculate the control amount U when the first local cost function is minimized 1 , to update the output of the power side subsystem;

[0053] In the synchronous sampling period k s , the three controllers on the power side, cold side, and hot side work together, and each controller receives the result ΔU of the last iteration of the other two controllers (p-1) , using ΔU (p-1) Calculate the control quantity U of each subsystem corresponding to the minimum local cost function in the new round of iteration 1 , U 2 and U 3 Each subsystem calculates the optimized control increment sequence and iterates and exchanges it, and the entire off-grid electric hydrogen coupled cooling, heating and power system reaches Nash equilibrium and obtains the Nash optimal solution U 1 , U 2 and U 3.

[0054] Compared with the prior art, the present invention has the following advantages:

[0055] The present invention first constructs an off-grid electric hydrogen coupled cooling, heating and power cogeneration system based on hydrogen-blended natural gas, and divides the off-grid electric hydrogen coupled cooling, heating and power cogeneration system into three subsystems: the electric side, the cold side and the hot side according to the differences in dynamic response characteristics. Then, a multi-time scale distributed predictive control method is adopted to design a local predictive controller to perform corresponding optimization control on the three subsystems: the electric side, the cold side and the hot side. In this way, the information interaction frequency between the three subsystems of the electric side, the cold side and the hot side can be designed according to the time scale of each variable according to the dynamic response time of different variables in the system, which better solves the problems of strong coupling of the system and large time scale span, and takes into account the tracking accuracy and response speed of the three subsystems of cooling, heating and power.

[0056] The present invention adopts multiple sampling cycles to perform real-time control of the three subsystems of the electric side, the cold side, and the hot side, which has the advantage of improving the rapidity of the system operation process. Taking into account the large difference in response time scales between the electric side and the temperature side (i.e., the cold side and the hot side), a multi-time scale information interaction mechanism is designed, which can speed up the response speed of the electric side while ensuring the control accuracy of the temperature side and save computing resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 It is a schematic diagram of the method flow of the present invention;

[0058] Figure 2 It is a framework diagram of an off-grid electric hydrogen coupled cooling, heating and power system in an embodiment;

[0059] Figure 3 It is a comparison diagram of the inertia time of each input and output in the embodiment;

[0060] Figure 4 It is a structural diagram of the distributed predictive control in the embodiment;

[0061] Figure 5 It is a multi-time scale information interaction diagram;

[0062] Figure 6 It is the control output curve diagram of MPC, DMPC and MTDMPC (method of the present invention) in the embodiment;

[0063] Figure 7 It is the control input curve diagram of MPC, DMPC and MTDMPC (method of the present invention) in the embodiment. DETAILED DESCRIPTION

[0064] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0065] Example

[0066] like Figure 1 As shown, a multi-time scale distributed predictive control method for an electric-hydrogen coupled integrated energy system comprises the following steps:

[0067] S1. Build an off-grid electric hydrogen-coupled combined heating and cooling power system based on hydrogen-blended natural gas;

[0068] S2. According to the differences in dynamic response characteristics, the off-grid electric hydrogen coupled combined heat and power system is divided into three subsystems: the power side, the cold side, and the heat side;

[0069] S3. Adopt multi-time scale distributed predictive control method, design local predictive controller, perform corresponding optimization control on the three subsystems of power side, cold side and hot side, and adjust the working state of off-grid electric hydrogen coupled combined heat and power system accordingly.

[0070] This embodiment applies the above solution, and the main contents are:

[0071] Step 1: Establish an off-grid electric hydrogen-coupled combined cooling, heating and power system based on hydrogen-blended natural gas to explore the dynamic characteristics of the system during operation;

[0072] Step 2: Divide the system into three subsystems: power side, cold side, and hot side according to the differences in the dynamic response characteristics of the system to achieve more refined distributed control;

[0073] Step 3: Multi-time scale distributed predictive control is adopted for the HCNG electric hydrogen coupled energy system, in which each local predictive controller will consider the performance indicator function of each subsystem and optimize the local performance indicator function respectively, so as to achieve the goal of improving the entire closed-loop system.

[0074] Step 1 specifically includes:

[0075] like Figure 2 As shown, the designed system is mainly composed of an adjustable heat recovery micro gas turbine, a double-effect lithium bromide absorption chiller, a solid oxide fuel cell, a cold storage tank, a battery, a hydrogen storage tank and other equipment, which can supply electricity, heat and cooling loads to building users at the same time.

[0076] Among them, HCNG passes through the separation device to supply natural gas and hydrogen to the micro-turbine and fuel cell respectively. In the micro-turbine, natural gas is directly sent to the combustion chamber, which produces high-temperature and high-pressure flue gas, driving the gas turbine and generator to convert chemical energy into electrical energy. At the same time, the waste heat flue gas is used for heating and cooling of the double-effect lithium bromide absorption refrigeration mechanism; in the fuel cell, hydrogen reacts with oxygen in the air to generate electricity, and the discharged high-temperature gas exchanges heat with the heating return water, and hot water is generated at the same time; the battery, as an energy storage unit, is only used for energy balance in the dynamic process, which can not only reduce the dependence of the cogeneration system on the battery, but also ensure the stability of user electricity consumption; the cold storage tank, as an energy storage unit, is responsible for adjusting the cooling flow rate to ensure stable cooling. However, to achieve real-time supply and demand balance of cold, heat and electricity, relying solely on the natural gas flow rate with a fixed hydrogen blending ratio cannot provide sufficient control flexibility. Therefore, this system uses the adjustment of the charging and discharging of the hydrogen storage tank to flexibly adjust the hydrogen flow rate at the inlet of the fuel cell, so that it can be used as another control degree of freedom to achieve stable energy supply of the system. On the user side, hot water is produced by the waste heat of the fuel cell exhaust and the lithium bromide unit.

[0077] The capacity configuration of each device in the system must ensure that it can fully meet the energy needs of users, and the minimum total cost of gas consumption cost, hydrogen cost and gas-hydrogen separation cost is adopted as the objective function. In this embodiment, the capacity configuration of each device is specifically: micro-gas turbine 80kW, double-effect lithium bromide unit 425kW, fuel cell 150kW, battery 507kWh, cold storage tank 1810kWh and hydrogen storage tank 500kWh.

[0078] Step 2 specifically includes:

[0079] According to the time scale diversity in the HCNG electric hydrogen coupled energy system, the system is decomposed into three subsystems: electricity, cooling and heat.

[0080] like Figure 3 As shown in Figure 2, in terms of electric power, the inertia time of the input variables is between 0 and 10.483 s, but μ reg 、m air The effect on electric energy is minimal. On the temperature side, the inertia time is distributed in the range of 35.73 to 736, and the difference is huge. The heating time constant is μ hgr The cooling time constant is nearly 21 times of that of the HCNG electric hydrogen coupling energy system. Therefore, the inertia time of the electric side and the temperature side differs by 3 to 368 times. The heating time constant and power supply time constant differ by 1000 orders of magnitude. Inside the temperature side, the difference in inertia time is nearly 21 times, and the system exhibits the characteristics of a large time scale span. From the dynamic characteristics analysis, it can be seen that μ reg The step change of heating temperature T hot The inertia time is greater than the cooling temperature Tcold The inertia time is nearly ten times longer. Obviously, μ hgr It is not suitable as the control quantity of the thermal subsystem. From the steady-state characteristic analysis, we know that μ reg The step change of cooling temperature T cold The steady-state changes have little effect, such as Figure 3 As shown, μ reg The heating temperature T is increased by a step of 20000s. ot The steady-state temperature changes by 0.295℃, which has an influence on the cooling temperature T cold The steady-state temperature of the cooling subsystem changes by only 0.0065°C, which is obviously not suitable as the control variable of the cooling subsystem. cold,CS Only affects the cooling temperature T cold , m air Only affects the heating temperature T ot Therefore, although the hot and cold subsystems have the same sampling period, it is still necessary to divide them into two subsystems for distributed control so that each subsystem can have a more refined control effect, which helps the system better cope with complex environmental changes and user needs and improves the system's response speed and accuracy.

[0081] In this embodiment, the volume flow rate of the hydrogen-mixed natural gas V mix and fuel cell inlet hydrogen flow rate As two control quantities of the power side subsystem, the bromine machine refrigerant valve opening μ hgr And the cold storage tank release flow m cold,CS As two control variables of the cold side subsystem, the opening of the micro-turbine regenerative valve μ reg and fuel cell inlet air flow rate m air As two control quantities of the hot side subsystem.

[0082] Step 3 specifically includes:

[0083] like Figure 5 As shown, this embodiment designs a multi-time scale distributed predictive control structure for the HCNG electric hydrogen coupled energy system.

[0084] Assume that the discrete sampling time of the model of electronic system #N is T N , the discrete sampling time of the model of the hot subsystem #H and the cold subsystem #C is T T , where T T It is T N The continuous state space model of the HCNG electric hydrogen coupling energy system is written in the partitioned discrete form as shown in equations (1), (2), and (3):

[0085]

[0086] Formula (1) is the state space expression of the power side subsystem, where k is the sampling step. Since the basic functions of MGT and SOFC are power supply, not heating, this subsystem controls the flow rate V of the mixed hydrogen natural gas. mix and the fuel cell inlet hydrogen flow To control the user end power N user . Power consumption N user At time interval T N sampling.

[0087]

[0088] Formula (2) is the state space expression of the cold side subsystem. The lithium bromide unit and the cold storage tank are responsible for the cooling task. Therefore, the opening of the refrigerant valve of the bromine unit μ hgr And the cold storage tank release flow m cold,CS It is placed in the thermal subsystem #C, which controls the opening of the bromine machine refrigerant valve μ hgr And the cold storage tank release flow m cold,CS To control the cooling temperature T cold . Cooling temperature T cold At time interval T T sampling.

[0089]

[0090] Formula (3) is the state space expression of the heat side subsystem. The high temperature flue gas of the micro-turbine and fuel cell undertakes the heating task by exchanging heat with other equipment. Therefore, the opening of the micro-turbine regenerative valve μ reg and fuel cell inlet air flow rate m air It is placed in the thermal subsystem #H, which controls the opening of the micro-turbine regenerative valve μ reg and fuel cell inlet air flow rate m air To control the heating temperature T hot . Heating temperature T hot At time interval T T sampling.

[0091] The power side controller takes into account the rapid tracking of user power load demand N user The control variable change of the power-side subsystem is small as the operation goal, and the local cost function (4) is used for subsystem #N;

[0092] The cold side controller takes into account the rapid tracking of user cooling load demand T cold The control variable change of the cold side subsystem is small as the operation goal, and the local cost function (7) is used for subsystem #C;

[0093] The hot side controller takes into account the fast tracking of user heat load demand T hotThe control variable change of the hot side subsystem is small as the operation goal, and the local cost function (8) is used for subsystem #H;

[0094]

[0095] In formula (4), N p is the prediction time domain, N c To control the time domain, Q 1,i is the error weighting coefficient, R 1,j To control the weighting coefficient, is the increment of system control action, ΔU 1,max , ΔU 1,min Respectively represent the upper and lower limits of the control action increment, U 1,max , U 1,min Respectively represent the upper and lower limits of the control action, where Y 1,p The system prediction output of subsystem 1 (i.e., the electrical side subsystem); is the system predicted output after feedback correction, which can be solved according to formula (5): is the expected trajectory of the system, which can be solved according to formula (6).

[0096] Y 1,p (k+i)=Y 1,m (k+i)+h 1 (Y 1 (k)-Y 1,m (k)) (5)

[0097] In formula (5), h is the correction coefficient, Y 1,m (k+i) is the multi-step prediction output of the electronic system.

[0098]

[0099] In formula (6), R 1 is the set value, ψ is the softening factor (0<ψ<1), which can make the system output Y 1 Smooth transition to the set value.

[0100]

[0101] In formula (7), Y 2,p is the system prediction output of subsystem 2 (i.e., the cold side subsystem), and its calculation method is similar to that of equation (4); ΔU 2 =[Δμ hgr ,Δm cold,CS ] is the input increment of subsystem 2.

[0102]

[0103] In formula (8), Y 3,pis the system predicted output of subsystem 3 (i.e., the hot side subsystem), and its calculation method is similar to that of equation (4); ΔU 3 =[Δμ reg ,Δm air ] is the input increment of subsystem 3.

[0104] In step 3, a multi-time scale distributed information interaction mechanism is adopted, which specifically includes:

[0105] Multi-time scale distributed information interaction mechanism Figure 5 As shown, the superscript p represents the number of iterations. The controller of the power-side subsystem receives the signal and updates the control input sequence at a faster rate, while the controllers of the cold-side and hot-side subsystems are updated at a slower rate. For better explanation, the time when the temperature-side (cold-side and hot-side) controller and the power-side controller are updated simultaneously is recorded as the synchronous sampling period k s , the time for updating the power side controller is recorded as the asynchronous sampling period k a .

[0106] In the asynchronous sampling period k a The power side controller will use the temperature side controller in the previous temperature side sampling cycle T T The calculated control quantity U 2 and U 3 Calculate the control quantity U when the local cost function (4) is minimized 1 , to update the output on the power side;

[0107] In the synchronous sampling period k s , the three controllers work together, which requires each controller to receive the results of the previous iteration of the other two controllers ΔU (p-1) , ΔU (p-1) It is used to calculate the control quantity U of each subsystem in the new round of iteration of the minimum local cost function (4), (7) and (8) 1 , U 2 and U 3 , each subsystem calculates the optimized control increment sequence and iterates and exchanges it, the whole system reaches Nash equilibrium, and the Nash optimal solution U of the system is obtained 1 , U 2 and U 3 .

[0108] Taking the information interaction of the electric side subsystem as an example, the step-by-step algorithm flow is as follows:

[0109] When t = k in synchronization period k s hour:

[0110] Step 1. Reset the number of iterations p = 1 and accept the initial value of the control variable change from the temperature side subsystem and And send the initial value of the electrical side to the temperature side subsystem

[0111] Step 2. Minimize the cost function (4) of subsystem #N and obtain And pass it to the temperature side subsystems #C, #H, setting p = p + 1;

[0112] Step 3. Repeat Step 1 until the conditions are met

[0113] Step 4. After the iteration stops, the sequence ΔU 1 (k)=[I,0,…,0]ΔU 1 (p) The first element in the previous moment U 1 (k-1) add to get U 1 (k) and executed into the controlled object.

[0114] When t=k is in asynchronous cycle k a hour:

[0115] Step 1. The electrical side receives the initial value of the control variable U from the temperature side subsystem 2 (k-1) and U 3 (k-1), U 2 (k-1) and U 3 (k-1) remains unchanged during the asynchronous cycle until the next synchronous cycle;

[0116] Step 2. Minimize the cost function (4) of subsystem #N and obtain ΔU 1 ;

[0117] Step 3. 1 (k)=[I,0,…,0]ΔU 1 The first element in the previous moment U 1 (k-1) add to get U 1 (k) and executed into the controlled object.

[0118] The design of multi-time scale information interaction can improve the response speed of the system's electrical side subsystem.

[0119] In order to verify the effectiveness of this solution, this embodiment verifies the control effect of the distributed intelligent predictive control of the present invention through a set of simulation comparison experiments, as follows: According to the "Gas Quality Requirements Entering Natural Gas Long-distance Pipelines" (GB / T37124-2018), the hydrogen blending ratio of natural gas pipelines is only allowed to be no more than 3%. The main operating parameters of the HCNG electrical hydrogen coupling energy system under typical working conditions are shown in Table 1. Based on the above analysis, this embodiment gives the control simulation results under cold, hot and electric load disturbances to verify the control effect of the proposed MTDMPC in the dynamic operation of the HCNG electrical hydrogen coupling energy system. The parameter settings of the control algorithm are shown in Table 2.

[0120] Table 1 Typical operating parameters of HCNG electric hydrogen coupled energy system

[0121]

[0122]

[0123] Table 2 Parameters of each controller

[0124]

[0125] This embodiment also simulates and compares the following two MPCs with the same control parameters:

[0126] a. Centralized MPC, which treats the system as a whole and transforms the control task into a centralized optimization problem. The sampling time of MPC is set to 3s;

[0127] b. Single time scale distributed model predictive control (DMPC), which consists of three cooperative MPCs for hot and cold electronic systems, with the same control structure as the MTDMPC described in the previous section. The three local MPCs have the same sampling time (T T =T N =3s).

[0128] c. Multi-time scale distributed model predictive control (MTDMPC), which consists of three cooperative MPCs for hot and cold electronic systems. The three local MPCs have different sampling times (T T =4s,T N =1s)

[0129] This simulation experiment focuses on verifying the load tracking accuracy and dynamic response speed of MTDMPC when the user side switches the set value. The initial state of the system output set value is [66.27, 7.66, 78.48], and it jumps to [86.27, 6.86, 78.98] at t = 500s, that is, the power supply increases by 20kW, the cold water outlet temperature set value decreases by 0.8℃, and the domestic hot water set value increases by 0.5℃. At t = 1400s, it jumps from [86.27, 6.86, 78.98] to [81.27, 7.36, 78.78], that is, the power supply decreases by 5kW, the cold water outlet temperature set value increases by 0.5℃, and the domestic hot water set value decreases by 0.2℃.

[0130] from Figure 6 and Figure 7 It can be seen that the centralized control scheme is no longer applicable to the HCNG electric hydrogen coupled energy system with multi-time scale properties and a large number of variables. By comparing the performance of the three algorithms, the necessity of distributed control is proved. Compared with DMPC, the MTDMPC algorithm has user The control of the cooling temperature T cold , heating temperature T hot The control also achieved good results. MTDMPC uses a shorter update cycle of the power-side sub-controller to better play the fast response characteristics of electricity. At the same time, by reducing the update and sampling frequency of the temperature-side sub-controller, the amount of calculation is reduced without affecting the control performance, so it can provide a faster adjustment speed than DMPC. The use of this algorithm can take on some of the anti-interference tasks of the battery unit, reduce the operating frequency of the battery, and extend the service life of the battery.

[0131] In summary, the present invention establishes an off-grid electric hydrogen coupled combined heat and power energy system based on hydrogen-blended natural gas, and through the simulation research and analysis of the system dynamic characteristics, it is found that the system has the characteristics of strong coupling and large time scale span. According to the differences in the dynamic response characteristics of the system, three subsystems, namely the electric side, the cold side and the hot side, are divided to achieve more refined control; the present invention proposes a multi-time scale distributed predictive control algorithm MTDMPC for the HCNG electric hydrogen coupled energy system. According to the dynamic response time of different variables in the system, multiple sampling cycles are used to control the subsystems in real time, which has the advantage of improving the rapidity of the system operation process, and taking into account the large difference in the response time scales of the electric side and the temperature side, a multi-time scale information interaction mechanism is designed to ensure the accuracy of the temperature side control while accelerating the response speed of the electric side and saving computing resources. The simulation experiment verifies that the proposed multi-time scale distributed predictive control algorithm can effectively improve the rapidity of the electric side tracking while meeting the tracking accuracy of the three types of cold, hot and electric loads of the system.

Claims

1. A multi-time scale distributed predictive control method for an electric-hydrogen coupled integrated energy system, characterized in that: The following steps are involved: S1. Build an off-grid electric hydrogen-coupled combined heating and cooling power system based on hydrogen-blended natural gas; S2. According to the differences in dynamic response characteristics, the off-grid electric hydrogen coupled combined heat and power system is divided into three subsystems: the power side, the cold side, and the heat side; S3. Adopt multi-time scale distributed predictive control method, design local predictive controller, perform corresponding optimization control on the three subsystems of power side, cold side and hot side, and adjust the working state of off-grid electric hydrogen coupled combined heat and power system accordingly.

2. A multi-time scale distributed predictive control method for an electric-hydrogen coupled integrated energy system according to claim 1, characterized in that: The off-grid electric hydrogen coupled cooling, heating and power system in step S1 includes a micro-turbine, a double-effect lithium bromide absorption refrigerator, a fuel cell, a cold storage tank, a battery, and a hydrogen storage tank.

3. The multi-time scale distributed predictive control method for an electric-hydrogen coupled integrated energy system according to claim 2, characterized in that: The operation process of the off-grid electric hydrogen coupled cooling, heating and power cogeneration system in step S1 is as follows: After HCNG passes through the separation device, natural gas and hydrogen are supplied to the micro-turbine and fuel cell respectively. In the micro-turbine, natural gas is directly sent to the combustion chamber, which produces high-temperature and high-pressure flue gas, driving the gas turbine and generator to convert chemical energy into electrical energy. At the same time, the waste heat flue gas is used for heating and cooling in the double-effect lithium bromide absorption refrigeration machine; In the fuel cell, hydrogen reacts with oxygen in the air to generate electricity, and the high-temperature gas discharged exchanges heat with the heating return water to generate hot water; Batteries are used as energy storage units for energy balance in dynamic processes; As an energy storage unit, the cold storage tank is responsible for regulating the cooling flow rate to ensure stable cooling supply; The hydrogen storage tank changes the hydrogen flow rate at the fuel cell inlet by adjusting the charge and discharge.

4. The multi-time scale distributed predictive control method for an electric-hydrogen coupled integrated energy system according to claim 2, characterized in that: In step S2, the control variables of the power-side subsystem are the volume flow rate of the hydrogen-mixed natural gas and the hydrogen flow rate at the fuel cell inlet, and the controlled variable is the user-side power; The control variables of the cold side subsystem are the opening of the refrigerant valve of the refrigerator and the cold release flow of the cold storage tank, and the controlled variable is the cooling temperature; The control variables of the heat side subsystem are the opening of the micro-turbine regenerative valve and the fuel cell inlet air flow rate; the controlled variable is the heating temperature.

5. The multi-time scale distributed predictive control method for an electric-hydrogen coupled integrated energy system according to claim 4, characterized in that: The state space expression of the electric side subsystem is: Among them, T N is the discrete sampling time of the electrical subsystem #N, V mix is the flow rate of hydrogen-mixed natural gas, is the hydrogen flow rate at the fuel cell inlet, and k is the sampling step length.

6. The multi-time scale distributed predictive control method for an electric-hydrogen coupled integrated energy system according to claim 5, characterized in that: The state space expression of the cold side subsystem is: Among them, T T is the discrete sampling time of the cold side subsystem #C, μ hgr is the opening of the refrigerant valve of the refrigerator, m cold,CS It is the cold release flow of the cold storage tank.

7. The multi-time scale distributed predictive control method for an electric-hydrogen coupled integrated energy system according to claim 6, characterized in that: The state space expression of the hot side subsystem is: Among them, T T is the discrete sampling time of the hot side subsystem #H, μ reg is the opening of the micro-turbine regenerative valve, m air is the fuel cell inlet air flow rate.

8. The multi-time scale distributed predictive control method for an electric-hydrogen coupled integrated energy system according to claim 7, characterized in that: Step S3 The process includes: S31, for the three subsystems of the power side, the cold side and the hot side, respectively design a corresponding power side controller, a cold side controller and a hot side controller, wherein the power side controller, the cold side controller and the hot side controller respectively adopt a first local cost function, a second local cost function and a third local cost function; S32, recording the time for simultaneously updating the cold side controller, the hot side controller and the electric side controller as a synchronous sampling period; The time when only the power-side controller is updated is recorded as the asynchronous sampling period; S33, adopting a multi-time scale distributed information interaction mechanism, in an asynchronous sampling period, the power-side controller takes minimization of the first local cost function as the control target and updates the control quantity of the power-side subsystem accordingly; In the synchronous sampling period, the electric side controller, the cold side controller and the hot side controller work together, taking the minimization of the corresponding local cost function as the control objective, and updating the control quantities of the three subsystems of the electric side, the cold side and the hot side accordingly.

9. The multi-time scale distributed predictive control method for an electric-hydrogen coupled integrated energy system according to claim 8, characterized in that: The first local cost function is specifically: Among them, N p is the prediction time domain of the electric side subsystem, N c is the control time domain of the electric side subsystem, Q 1,i is the error weighting coefficient of the electric side subsystem, R 1,j is the control weighting coefficient of the electric side subsystem, is the control action increment of the electric side subsystem, ΔU 1,max , ΔU 1,min They represent the upper and lower limits of the control action increment of the electric side subsystem, U 1,max , U 1,min They represent the upper and lower limits of the control action of the electric side subsystem, Y 1,p System prediction output for the electric side subsystem; is the predicted output of the electric side subsystem after feedback correction, is the expected trajectory of the electric side subsystem, h is the correction coefficient, Y 1,m (k+i) is the multi-step prediction output of the electric side subsystem, R1 is the set value, ψ is the softening factor, 0<ψ<1, which is used to make the output Y1 of the electric side subsystem smoothly transition to the set value; The second local cost function is specifically: Among them, Y 2,p is the system predicted output of the cold side subsystem, Y 2,r is the expected trajectory of the cold side subsystem, ΔU2=[Δμ hgr ,Δm cold,CS ] is the control action increment of the cold side subsystem, Q 2,i is the error weighting coefficient of the cold side subsystem, R 2,j is the control weighting coefficient of the cold side subsystem, ΔU 2,max , ΔU 2,min They represent the upper and lower limits of the control action increment of the cold side subsystem, U 2,max , U 2,min They represent the upper and lower limits of the control action of the cold-side subsystem respectively; The third local cost function is specifically: Among them, Y 3,p is the system predicted output of the hot side subsystem, Y 3,r is the expected trajectory of the hot side subsystem, ΔU3=[Δμ reg ,Δm air ] is the control action increment of the hot side subsystem, Q 3,i is the error weighting coefficient of the hot side subsystem, R 3,j is the control weighting coefficient of the hot side subsystem, ΔU 3,max , ΔU 3,min They represent the upper and lower limits of the control action increment of the hot side subsystem, U 3,max , U 3,min They represent the upper and lower limits of the control action of the hot side subsystem respectively.

10. The multi-time scale distributed predictive control method for an electric-hydrogen coupled integrated energy system according to claim 9, characterized in that: The specific process of step S33 is as follows: In the asynchronous sampling period k a The electric side controller uses the cold side and hot side controllers in the previous temperature side sampling cycle T T The calculated control quantities U2 and U3 are used to calculate the control quantity U1 when the first local cost function is minimized, so as to update the output of the power side subsystem; In the synchronous sampling period k3, the three controllers on the power side, cold side, and hot side work together, and each controller receives the result ΔU of the last iteration of the other two controllers. (p-1) , using ΔU (p-1) The control quantities U1, U2 and U3 of each subsystem corresponding to the minimum local cost function in the new round of iteration are calculated. Each subsystem calculates the optimized control increment sequence and exchanges them iteratively. The entire off-grid electric hydrogen-coupled combined heat and power system reaches Nash equilibrium and obtains the Nash optimal solutions U1, U2 and U3.

Images