Due to the increasing demand for the sustainability of energy use, the use of renewable energy has recently attracted interest [1]. Solar thermal power generation has attracted widespread attention because of its energy storage, peak adjustment, and continuous power generation, excellent power quality, and direct access to the grid [2]. At present, the commercial solar thermal power generation projects that have been built in the world are mainly parabolic trough and tower. The two technologies are relatively mature, but the construction cost is difficult. Linear Fresnel technology is relatively new, with simple structure, low construction and operating costs, strong wind and sand resistance, and obvious advantages in Western China. However, due to its long transmission pipeline and direct exposure to the air, the linear Fresnel type solar thermal power generation projects that have been built in the world all use water or thermal oil as the working medium, and the heat storage efficiency is low [3]. At the end of 2019, China completed the world’s first commercial project of molten salt linear Fresnel solar thermal power generation-Dunhuang Fresnel solar thermal project, using binary molten salt as the direct heat transfer and storage medium, the work temperature of molten salt is between 290°C–580°C [4]. Unlike systems that use heat transfer oil as the working fluid, molten salt as a heat transfer fluid allows the system to have a higher heat transfer temperature while reducing the system pressure, resulting in higher thermodynamic efficiency and safety [5].

Hybrid energy system management has always been a major issue in the field of energy engineering. In the field of hybrid batteries, literature [6] proposed an energy management strategy based on adaptive rules to extend the battery life of solar wireless sensor nodes. Literature [7] used a multi-objective optimization algorithm to solve the problem of using retired electric vehicle batteries in the PV-hydrogen-REVB hybrid energy system. Literature [8] proposed an independent hybrid microgrid management system with broad application prospects. Literature [9] proposes to use supercapacitor-battery hybrid energy storage system to extend the cycle life of the battery. The solar thermal power generation system is also a comprehensive energy management system. The difference is that there are very complex energy coupling and conversion relationships in the solar thermal power generation system.

At present, the solar thermal power station with heat storage system that uses molten salt as the heat transfer medium is the main development direction at present, and it is necessary to study the system. Literature [10] combined the thermo-hydrodynamic model of the trough collector with the model of the traditional steam power plant, and established a unified model of the solar power system, but lacked systematic analysis of complex working conditions. Literature [11] verified through experimental tests that compared with other concentrated solar power (CSP) systems, linear Fresnel technology is more suitable for direct molten salt (DMS) heat storage technology, but the current research on CSP systems using DMS is mainly on a certain local module or a certain transient dynamic modeling. Literature [12] used the one-dimensional steady-state distributed parameter model of the collector and uses an object-oriented method to establish the transient model of the parabolic trough collector with molten salt as the heat transfer fluid. Literature [13] established a dynamic model of the CSP system steam power generation system through the lumped parameter method, and simulated the interference that solar thermal electric plants might encounter under variable load conditions. Literature [14] proposed an economic model of solar trough solar thermal power station, which improved the efficiency of solar energy utilization. Literature [15] developed a simulation tool for solar thermal power generation to simulate and analyze the energy conversion process of solar thermal cycle and power cycle respectively. However, the solar thermal power generation system will have different coupling operation modes under different working conditions. In order to achieve the energy conversion and mode switching of the system at the same time, it is necessary to use the modeling method of the hybrid system to realize the hybrid control of the system. Table 1 shows the comparison of related literature and the research content of this article.

Petri net is an important modeling and analysis tool for describing hybrid systems. It can simultaneously describe the control of discrete processes on continuous variables and the dynamic processes of continuous variables. Mixed state Petri nets are widely used in the modeling, analysis and design of temperature control and working condition switching systems. Literature [16] proposed a hybrid petri net model (hybrid state Petri net, HSPN) for process control hybrid system modeling, and designed a hybrid controller based on the HSPN model. Literature [17] established the switching criterion and switching strategy for the large-scale heating process of the electric heating furnace, and realized the accurate and stable switching of the system by using the Petri net model. Literature [18] proposed a hybrid Petri net method to model the system for the scheduling problem of crude oil with different melting points in the refinery, and determine the schedule ability condition by analyzing the dynamic behavior of the network model. Literature [19] proposed a hybrid random timing Petri net for a type of hybrid system, which can simultaneously represent a hybrid system with discrete, continuous, conflict, delay and random characteristics. By establishing an equivalent model of HPN, the advantages of HPN in describing hybrid systems are verified. Literature [20] provided a new hybrid control method based on Petri nets, which combines the existing discrete event system theory and continuous system dynamic theory to model, analyze and synthesize hybrid systems. CSP system has both continuous variables and discrete processes during operation, which is a typical hybrid system [21]. The dynamic equation of the energy conversion process and the logic switching criterion of the mode switching process are established respectively. Based on this, the ex-tended differential Petri net model of the CSP system is established, and the state evolution behavior of the system is analyzed, so as to provide practical engineering for the system using molten salt for heat storage. The above operating program provides a theoretical basis.

The heat collecting field focuses the solar energy and heats the molten salt flowing through the heat collecting pipe to convert the solar energy into heat energy. The heat energy absorbed by the heat collecting field is shown in Eq. (1) [22]. (1)q˙abs=IGηoptαabsτenvwhere, I is the direct solar radiation intensity, G is the aperture area, ηopt is the optical efficiency, αabs is the absorption rate of the absorption tube, and τenv is the transmittance of the glass sealed tube.

The most significant transient effect in the collector field is the thermal mass of the HTF in the collector tube and pipeline. Therefore, the energy change of the thermal mass needs to be considered when deriving the energy balance equation. Assuming that the incompressible HTF passes through a constant volume heat collection element, its energy balance includes inlet energy q˙in,sf, outlet energy q˙out,sf, absorbed energy q˙abs and an internal energy item ∂U∂t. The total energy balance in the control volume is: (2)q˙in,sf+q˙abs=∂U∂t+q˙out,sf

The heat of import and export can be expressed as: (3)q˙in,sf−q˙out,sf=chtfm˙out,sf(Tin,sf−Tout,sf)→

The internal energy term represents the energy change of thermal mass, which can be expressed as: (4)∂U∂t=(mhtfchtf+(mc)balL)∂T∂twhere, m˙ is the mass flow rate of the HTF, Tin,sf is the molten salt temperature at the inlet of the collector, Tout is the molten salt temperature at the outlet of the collector, mhtf is the mass of the HTF in the pipeline, (mc)bal is the thermal inertia term of the collector component, and L is the temperature of the collector component length.

Substituting Eqs. (3) and (4) into Eq. (2), the partial differential equation of the molten salt temperature at the outlet of the collector field is obtained as shown in Eq. (5). (5)∂T∂t=chtfm˙out,sf(Tin−Tout)+q˙abs(mhtfchtf+(mc)balL)

The heat storage system adopts a double tank model, and uses molten salt sensible heat to store heat. The low-temperature heat storage tank provides liquid molten salt for system operation, and the high-temperature heat storage tank stores heat to ensure the stable power generation of the system. The low-temperature heat storage tank and the high-temperature heat storage tank operate independently, but the same model can be used to simulate the energy conversion process [9].

The mass conservation equation of the thermal storage salt tank is as follows: (6)mhot=m0,hot+m˙in,hotΔt−m˙out,hotΔtwhere, m0,hot is the initial value of the molten salt mass in the high-temperature thermal storage tank, m˙in,hot is the mass flow of molten salt at the inlet of the high-temperature thermal storage tank, and m˙out,hotis the mass flow of molten salt at the outlet of the high-temperature thermal storage tank.

In order to prevent the molten salt in the molten salt tank from condensing, the electric accompanying heating method is used to provide supplementary heat to the HTF in the low-temperature heat storage tank. The required heat is calculated based on the volume and temperature of the HTF at the end of the time step, as shown in Eq. (7) shown. (7)q˙fp,cold=chtfρhtfπr2hcold(Tset,cold−Tcold)

The heat loss model of the low-temperature storage tank is shown in Eq. (8). (8)q˙hl,cold=UA(Tcold−Ta)where, q˙ is the heat transfer rate, c is the specific heat capacity, ρ is the density, r is the radius of the thermal storage tank, hcold is the liquid level of the low-temperature thermal storage tank, Tset,cold is the starting temperature of the low-temperature thermal storage tank electric heating device, and Tcold is the temperature in the low-temperature thermal storage tank, molten salt temperature Ta is the ambient temperature, and UA is the heat loss coefficient. The subscript htf means heat transfer fluid, cold means low temperature storage tank, and hl means heat loss.

The power cycle consists of a dual-tank heat storage subsystem, a steam generation subsystem, and a power generation subsystem. The conversion model from thermal energy to electrical energy is established according to the basic heat transfer equation on the hot side of the heat exchanger, and the thermal energy provided by the high-temperature heat storage tank is shown in Eq. (9). (9)q˙therm=m˙out,hotchtf(Tout,hot−Tin,cold)where, m˙out,hot=m˙in,cold.

Assuming that the system provides the same power output during the simulation time, the rated power generation of the system at the design point is P = 50 MW, the conversion efficiency of the known rated cycle is ηdes=0.397, and the conversion coefficient from gross power to net power is ηelc=0.9, then the heat storage system needs to provide The thermal energy is shown in Eq. (10). (10)q˙therm=P/ηelcηdes

Definition 1. The extended differential Petri net is an octet [24], EDPNs={P,T,A,W,h,f,M0,J}.

Definition 2. The conditions under which changes occur. In Petri net, the library represents the state of the system, and the change of the system state is described by transition. The Petri net’s network structure is static, and its dynamic characteristics are through the continuous occurrence of discrete events, so that the enabling conditions of change are stimulated, and the conditions for the occurrence of changes are as follows:

If f(tj)=D, then is enabled at time t if and only if pi∈tj⇒h(pi,tj)(M(t)) is true;

Definition 3. Evolution rule. The enablement of change leads to the continuous change of the state of the place, which realizes the evolution of Petri net. The transition tj is enabled at time t, the excitation speed is vj(t), and the integration step is dj. Its excitation will cause the logo to be converted according to the following rules:

If f(tj)=D and J(tj)=τj, then there is mi(t+τj)={mi(t)−Pre(pi,tj),∀pi∈∘tjmi(t)+Post(pi,tj),∀pi∈tj∘mi(t),other

Definition 4. Net activity [25]. If for any reachable mark M∈[M0⟩, there is a reachable mark of M′∈[M⟩, and t has the right to occur in M′, then t is alive, which also includes control by the operator to trigger changes, which means control for the control system The operability. If all changes in EDPNs are alive, EDPNs are alive.

According to the differential equations of continuous variables and the switching conditions of discrete states, the Petri net model of the solar thermal power generation system is established as shown in Fig. 5. Among them, p₃, p₆, p₇, p₈, p₁₃, p₁₆ are discrete places, and p₁, p₂, p₄, p₅, p₉, p₁₀, p₁₁, p₁₂, p₁₄, p₁₅, p₁₇ and p₁₈ are continuous places. The meaning of the place is shown in Table 2.

After the system is initialized, it enters the waiting state. When is satisfied q˙abs>0∩hcold>hmin∩Tout,sf≤Tdes,sf, the transition t₁ is enabled, and the low-temperature molten salt pump is turned on to provide low-temperature molten salt to the heat col-lection field. At this time m₃=1, t₃ is activated, and the identification value in p₅ is calculate. If m5<425, t₆ is activated and the bypass valve is opened. At this time, m₆=1, t₉ is activated, and the low-temperature molten salt entering the heat collection tube is not heated to the desired temperature, so it circulates back to the low-temperature molten salt tank to achieve the effect of preheating the heat collection field. When the switching conditions q˙abs>0∩hcold>hmin∩Tout,sf>Tdes,sf are met, the transition t₅ is enabled, and the main circuit valve is opened. At this time, m₇=1, t₈ is excited, and the excitation speed, the low-temperature molten salt flows through the heat collecting field, and after being heated to the desired temperature, it is stored as a high-temperature molten salt in the high temperature heat storage tank. When the switching conditions q˙hot>q˙dem∩Tout,hot≤Tdes,hot are met, the transition t₁₀ is enabled, and the high-temperature molten salt pump is turned on. At this time, m₁₁=1, t₁₄ is ex-cited, and the heat storage system provides high-temperature molten salt to meet the power generation requirements of the system. When the switching conditions Tcold<Tset,cold are met, the transition t₄ is enabled, and the external power block of the low-temperature heat storage tank is turned on. At this time m₈=1, t₇ is activated; when the switching conditions Thot<Tset,hot are met, the transition t₁₀ is enabled and the external power block of the high-temperature heat storage tank is activated. Turn on, at this time m₁₁=1, t₁₃ is activated, and the molten salt in the heat storage tank is protected against condensation. Because the temperature of the high and low temperature storage tank is higher than the ambient temperature, there is always heat loss. The speed and weight of changes are shown in Table 3.

Through comparison with experimental data, the validity of the Petri net model is verified. Fig. 6 shows the direct solar irradiance received by the solar electric field within a week. Figs. 7 and 8 respectively represent the loop outlet temperature and HTF mass flow rate and relative deviation calculated by the simulation tool and experimental data during the whole week. It can be seen from these charts that the simulation results of the model and the experimental data generally match well. Due to the large fluctuation of DNI on No. 9.17, the relative deviation of temperature reached 10%, but the deviation of cumulative net heat transfer results was less than 1%. It is acceptable for this study. The Petri net model accurately captures the system dynamics in different operating modes, and also ensures continuous conversion between operating modes.

The Petri net model simulates the thermal energy transfer and mode switching process of the entire factory by considering the solar heat and power cycle of the entire system model. Two days with different irradiance levels were used to simulate and analyze the energy transfer and mode switching of the system, as shown in Figs. 9 and 10. The simulation was carried out for a full 24 h, which is the complete operating cycle of the solar field.

As shown in Fig. 9, at the beginning of each day, the low-temperature molten salt pump is turned on to output low-temperature molten salt to the heat collection field, and the HTF circulates back to the low-temperature heat storage tank through the heat collection field to preheat the heat collection field. The thermal field gradually warms, the inlet temperature of the collector field rises and the liquid level of the high-temperature storage tank does not change. During this period, the main reflector continues to track the sun (defocus rate = 0%), as shown in Fig. 9 from 6:00 to 17:00 this time. When the allowable inlet temperature of the high-temperature heat storage tank is reached (in this example, it is about 425°C), HTF is circulated to the high-temperature heat storage tank and stores heat, as shown in Fig. 9 from 7:00 to 16:00. As long as the HTF temperature output by the heat collection field reaches the set temperature and the high-temperature salt storage tank is not full, the HTF is completely guided to the high-temperature heat storage tank. Otherwise, the HTF is recycled into the low-temperature heat storage tank, as shown in the two time periods of 6:00 to 7:00 and 16:00 to 17:00 in Fig. 9. The outlet temperature of the heat collection field is maintained at its rated level by controlling the mass flow rate of the HTF. If the DNI exceeds the absorption capacity of the heat collection field, it will be defocused by the main reflector. At the end of the day, when the available DNI is not enough to maintain the minimum operating temperature at the minimum mass flow rate, the HTF mass flow rate is gradually reduced to the minimum value until the DNI is 0, the main mirror is defocused, and the heat collection field begins to slowly cool. During the period from 08:00 to 01:00 the next day, the heat in the high-temperature heat storage tank can meet the output demand of the steam turbine, and the high-temperature molten salt pump is turned on to provide high-temperature molten salt to the steam generator and generate steam to drive the steam turbine unit For power generation, the low-temperature molten salt after heat exchange is recycled to the low-temperature molten salt tank as a reserve of molten salt at the entrance of the heat collection field.

In order to better evaluate the performance of the solar thermal power generation system, it is necessary to calculate the thermal performance of the system under annual working conditions. Select the local annual irradiance in 2018 as the input data, and the monthly statistics of solar resources are shown in Fig. 11.

Using the typical annual radiation data, the simulated monthly average power generation and cumulative value of the molten salt direct heat storage linear Fresnel power station are shown in Fig. 12, and the total annual power generation is 287.82 GWh. The average annual power generation efficiency is one of the important factors for evaluating the economic performance of solar thermal power plants. The annual power generation efficiency is the ratio of the power generation capacity of the power station in one year to the annual accumulated solar radiation collected, namely: (11)η=∑PeΔt∑IAΔt×100%

For solar power plants, the two parts involved in energy conversion are the heat collection subsystem and the power generation subsystem. The heat collection sub-system focuses the solar energy on the surface of the heat collection tube through the main reflector, and then heats the heat transfer flowing through the heat collection tube. The medium realizes the conversion of light energy to heat energy; the power generation subsystem transfers the heat of the high-temperature molten salt to the water vapor through the steam generator, and finally uses the superheated steam to drive the steam turbine unit to realize the conversion of heat energy to electric energy. Using the annual radiation data, the calculated energy absorbed by each energy con-version process is shown in Fig. 13. It can be seen that there is a large amount of energy loss during the photothermal conversion process, and its annual average photo-thermal conversion efficiency is 39.22%. Secondly, the heat loss of the steam turbine generator unit in the thermoelectric conversion process is more, and the average annual thermoelectric conversion efficiency is 40.97%. According to formula 4.8, the thermo electric conversion efficiency of the entire solar thermal power generation system is 16.04%. Therefore, improving the photothermal conversion efficiency and thermoelectric conversion efficiency of the system is an effective measure to improve the photoelectric conversion efficiency of the system, effectively use solar resources, and reduce the cost of electricity. Increasing the heat source temperature can improve the thermo-electric conversion efficiency, but an excessively high operating temperature will reduce the safety and stability of the system. Therefore, improving the light-to-heat conversion efficiency of the collector at an optimal temperature is the main break-through to reduce the cost of solar thermal power stations.