A hierarchical distributed control method and system for a multi-energy ship hybrid power system

Abstract

The present invention discloses a hierarchical distributed control method for a multi-energy ship hybrid power system, comprising: predicting the ship's required power according to a ship load prediction model; calculating the reference power of batteries, diesel engines, motors, and generator sets; calculating the SOC trajectory of the batteries according to the reference power of the batteries; tracking the SOC trajectory to calculate new reference power of batteries, diesel engines, motors, and generator sets; calculating the reference control voltage compensation ΔU of each power source according to the deviation between the output voltage of the power source and the expected voltage; dynamically calculating the droop coefficient according to the new reference power of batteries, diesel engines, motors, and generator sets; and calculating the reference control voltage V of each power source through a droop controller. ref ; For reference control voltage V ref The reference control voltage compensation value ΔU is superimposed and processed as the final control voltage output to achieve hierarchical distributed control. The present invention can meet the real-time optimization and precise power distribution of multi-energy hybrid power systems.

Description

Technical Field

[0001] The present invention belongs to the field of ship hybrid power systems, and in particular relates to a hierarchical distributed control method and system for a multi-energy ship hybrid power system. Background Art

[0002] Multi-energy hybrid ships can achieve energy conservation and emission reduction in three ways: first, by integrating renewable energy and energy storage devices to reduce fossil energy consumption; second, by reducing the selected power of the main engine to improve the problem of long-term low-load operation of the main engine during navigation; and third, by adopting intelligent energy management strategies to improve the energy efficiency of the hybrid power system. With the development of green and intelligent ships, new energy sources such as wind power, photovoltaic cells, and fuel cells, as well as energy storage devices such as batteries and supercapacitors, are gradually being used on ships, making the structure and control of marine hybrid power systems increasingly complex. Energy management systems not only manage, distribute, and control the electrical energy of the ship's power system, but also involve the control of specific power equipment and power devices to achieve optimal power distribution among various power sources to meet the ship's dynamic, economic, and emission requirements. The safe and efficient operation of hybrid ships depends on the hybrid power system and energy management system.

[0003] The energy management system with a hierarchical distributed control structure combines the advantages of centralized control and decentralized control, with the advantages of a simple structure, clear responsibilities, and a clear division of labor. Although the hierarchical distributed control structure has been applied to some urban power grid systems, it still has many shortcomings in the marine industry. It has not fully integrated the actual application scenarios of ships, and there has been no targeted research and design of a highly feasible hierarchical distributed control structure. The existing ship energy management control structure has two main shortcomings:

[0004] (1) The energy management layer lacks research and design of real-time optimization algorithms with higher engineering application value. Existing scholars have proposed many energy management methods, but they mainly focus on designing global optimization algorithms in scenarios where ship operating conditions are known. There is a lack of design of real-time optimization energy management strategies that combine the power control layer and the local control layer, and lack the rationality of engineering application.

[0005] (2) The local control layer usually adopts a fixed droop coefficient control method, which cannot meet the application requirements of the optimization energy management strategy. Since the optimization result of the optimization energy management strategy is the reference distribution power of multiple distributed power sources, the distribution ratio is often nonlinear and non-fixed. The traditional droop control method mainly designs a fixed droop coefficient according to the capacity size, that is, corresponds to a fixed power distribution ratio. Therefore, the traditional droop control method obviously cannot meet the requirements of the optimization energy management strategy. Summary of the Invention

[0006] The purpose of the present invention is to provide a hierarchical distributed control method and system for a multi-energy ship hybrid power system, which can meet the real-time optimization and precise power distribution of the multi-energy hybrid power system.

[0007] To solve the above technical problems, the technical solution of the present invention is: a hierarchical distributed control method for a multi-energy ship hybrid power system, comprising the following steps:

[0008] S1. Predict the ship's required power from time k+1 to time k+N based on the ship load prediction model;

[0009] S2. Using a forward dynamic programming algorithm, with minimum equivalent fuel consumption as the optimization goal, calculate the reference charge and discharge power of the battery, the reference output power of the diesel engine, the reference output power of the motor, and the reference output power of the generator set from time k+1 to time k+N;

[0010] S3. Calculate the SOC trajectory of the battery from time k+1 to time k+N based on the reference charge and discharge power of the battery from time k+1 to time k+N;

[0011] S4. Track the SOC trajectory through a model predictive controller, with minimizing the tracking error as the optimization goal, and calculate the new reference charge and discharge power of the battery, the reference output power of the diesel engine, the reference output power of the motor, and the reference output power of the generator set from time k+1 to time k+N;

[0012] S5. Calculate the reference control voltage compensation ΔU of each power source based on the consistency algorithm according to the deviation between the output voltage of the power source and the expected voltage;

[0013] S6. Dynamically calculate the droop coefficient based on the new reference charge and discharge power of the battery, the reference output power of the diesel engine, the reference output power of the motor, and the reference output power of the generator set; calculate the reference control voltage V of each power source through the droop controller. ref ;

[0014] S7, reference control voltage V ref The reference control voltage compensation ΔU is superimposed as the final control voltage output to control the output voltage of each power source and realize hierarchical distributed control.

[0015] The model predictive controller in S4 tracks the SOC trajectory through the model predictive control method. The model predictive control method tracks the SOC trajectory based on the established hybrid system state equation, takes minimizing the tracking error as the optimization goal, and calculates the new battery reference charge and discharge power, diesel engine reference output power, motor reference output power and generator set reference output power from time k+1 to k+N.

[0016] The hybrid system state equation is established based on the relationship between the battery's SOC trajectory and the charge and discharge power and time; the SOC trajectory is the relationship between the battery's remaining power and time.

[0017] The forward dynamic programming algorithm in S2 is based on the current system SOC state and the predicted ship power demand. According to the characteristics of the battery, the possible output power of the battery from time k+1 to k+N is discretized, and the optimal solution is found through traversal search. The optimization goal of the optimization process is to minimize the equivalent fuel consumption.

[0018] The load prediction in the ship load prediction model in S1 is implemented based on a machine learning intelligent algorithm; its prediction time domain is set according to the accuracy of the ship operating condition prediction.

[0019] S5 is based on the deviation between the ship bus voltage and the expected value. By adjusting and compensating the voltage output by the local controller, the ship grid bus voltage is brought close to the expected value. The consistency algorithm communicates and exchanges information based on the output voltage of each power source to obtain the reference control voltage compensation. The reference control voltage compensation is obtained through PI control and its calculation method is:

[0020] ΔU i =k p (V0-V ave )+k i ∫(V0-V ave )dt

[0021] Among them, ΔU i is the reference control voltage compensation output by the distributed controller; V i is the output voltage of the i-th power source; k p and k i All are PI controller parameters; V ave is the average voltage of each power source obtained based on the consistency algorithm; V0 is the expected value of the ship power grid bus voltage.

[0022] V ave The calculation method is:

[0023] X[k+1]=D·X[k]

[0024] Where X[k+1] is the output voltage matrix of the discretized distributed power source at the kth moment; D is the consistency matrix determined according to the topology of the control object based on the consistency theory.

[0025] The droop controller calculates the reference control voltage V of each power source through the droop control method based on dynamic virtual impedance. refThe droop control method is based on the characteristic that the output voltage of the power source decreases as the power increases. It is assumed that the droop coefficient of the droop controller is constant. The principle of droop control is:

[0026] V ref_i (t) = V0 - k i I i (t)

[0027] Where V ref_i (t) is the reference control voltage of the i-th power source at time t; i is the droop coefficient of the i-th power source.

[0028] The droop control method based on dynamic virtual impedance is an improvement on the traditional droop control method. The improvement is to dynamically design the droop control coefficient. The real-time droop coefficient is calculated based on the new reference charge and discharge power of the battery, the reference output power of the diesel engine, the reference output power of the motor, and the reference output power of the generator set. The calculation method of the droop coefficient is established based on the characteristics of droop control, that is, the power size is inversely proportional to the droop coefficient. The real-time droop coefficient calculation formula is:

[0029]

[0030] p motor is the new motor reference output power; p battery is the reference charge and discharge power of a new battery. battery >0, discharge, when p battery <0, then charge.

[0031] A multi-energy ship hybrid power system hierarchical distributed control system is also provided, comprising an energy management layer, a power control layer and a local control layer arranged in sequence; wherein,

[0032] The energy management layer is used to predict the ship's power demand from time k+1 to k+N based on the ship load prediction model; using a forward dynamic programming algorithm, with minimum equivalent fuel consumption as the optimization goal, calculate the reference charge and discharge power of the battery, the reference output power of the diesel engine, the reference output power of the motor, and the reference output power of the generator set from time k+1 to k+N; based on the reference charge and discharge power of the battery from time k+1 to k+N, calculate the SOC trajectory of the battery from time k+1 to k+N; using a model predictive controller to track the SOC trajectory, with minimum tracking error as the optimization goal, calculate the new reference charge and discharge power of the battery, the reference output power of the diesel engine, the reference output power of the motor, and the reference output power of the generator set from time k+1 to k+N;

[0033] The power control layer is used to calculate the reference control voltage compensation ΔU of each power source based on the deviation between the output voltage of the power source and the expected voltage using a consistency algorithm;

[0034] The local control layer is used to dynamically calculate the droop coefficient based on the new reference charge and discharge power of the battery, the reference output power of the diesel engine, the reference output power of the motor, and the reference output power of the generator set; the reference control voltage V of each power source is calculated by the droop controller. ref ; For reference control voltage V ref The reference control voltage compensation ΔU is superimposed as the final control voltage output to control the output voltage of each power source and realize hierarchical distributed control.

[0035] Compared with the prior art, the present invention has the following beneficial effects:

[0036] First, by considering the functional characteristics and real-time requirements of different control layers, a rational hierarchical distributed control structure was designed, integrating the robustness of MPC control with the optimization performance of a dynamic programming algorithm. The designed control structure is simple, highly feasible, and has good real-time performance, with excellent theoretical and engineering application value.

[0037] Secondly, it fully combines the advantages of the forward dynamic programming (FDP) algorithm and the model predictive control (MPC) method. The FDP algorithm obtains the reference trajectory for optimal energy saving by predicting the operating conditions, while the MPC method provides a highly robust reference power result by following.

[0038] Third, a dynamic virtual impedance method is proposed to precisely control the desired power distribution. This method is not limited to the hybrid system topology discussed above and can be extended to hybrid systems with more distributed power sources. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 A schematic diagram of a flow chart of an embodiment of the present invention;

[0040] Figure 2 Schematic diagram of the structure of an embodiment of the present invention;

[0041] Figure 3 Schematic diagram of the principle of an embodiment of the present invention;

[0042] FIG4( a ) is a schematic diagram of the energy flow topology structure of a generator supplying energy to drive a motor according to an embodiment of the present invention;

[0043] FIG4( b ) is a schematic diagram of the energy flow topology structure of a generator supplying energy to drive the motor and charging the battery in an embodiment of the present invention;

[0044] FIG4( c ) is a schematic diagram of the energy flow topology structure of a generator and a battery-powered drive motor according to an embodiment of the present invention;

[0045] FIG4( d ) is a schematic diagram of the energy flow topology structure of two generators supplying energy to drive the motor according to an embodiment of the present invention;

[0046] FIG4( e ) is a schematic diagram of the energy flow topology structure of two generators supplying energy to drive the motor and charging the battery in an embodiment of the present invention;

[0047] FIG4( f ) is a schematic diagram of the energy flow topology structure of two generators and a battery-powered drive motor in an embodiment of the present invention;

[0048] In the figure, 1-first ship generator set, 2-second ship generator set, 3-battery, 5-first converter device, 6-second converter device, 7-third converter device, 8-fourth converter device, 9-ship grid bus, 10-ship motor, 11-ship diesel engine, 12-gearbox, 13-ship propeller. DETAILED DESCRIPTION

[0049] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely for the purpose of explaining the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below may be combined with each other as long as they do not conflict with each other.

[0050] The topology of the multi-energy ship hybrid power system shown in FIG4 is used as the control object of the hierarchical distributed control method.

[0051] According to the hierarchical distributed control method, the implementation process is shown in Figure 1 The specific implementation steps are as follows:

[0052] Step 1: Predict the ship's power demand at future times based on the ship load prediction model;

[0053] Step 2: The forward dynamic programming algorithm in the energy management layer calculates the optimal diesel engine output power, ship motor output power, battery charge and discharge power, and generator set output power at the future time based on the current battery SOC state, the power demand at the future time, and the optimization goal of minimizing equivalent fuel consumption.

[0054] Step 3: Based on the obtained battery future discharge trajectory and the battery's own characteristics, the predicted SOC trajectory at the future moment is calculated and used as the reference input of the model predictive controller.

[0055] Step 4: The model predictive controller in the energy management layer tracks the battery SOC reference trajectory obtained in Step 3 and solves for the battery reference power again, ensuring that the SOC change trajectory corresponding to the state equation in the model predictive controller is consistent with the reference trajectory. Based on the new battery reference power, the optimal generator set output power and ship motor output power are determined at future times. The ship diesel engine power remains unchanged based on the solution obtained by the FDP algorithm.

[0056] Step 5: The energy management layer outputs the optimal diesel engine output power, ship motor output power, battery charge and discharge power, and generator set output power at the future moment. The local control layer receives the above reference power results and calculates the real-time droop coefficients of different power sources based on the dynamic virtual droop control method, as follows:

[0057]

[0058]

[0059]

[0060] P motor =P generator1 +P generator2 +P battery

[0061] P demand_pred =P motor +P diesel

[0062] Among them, k generator1 , k generator2 are the current droop coefficients of the two generator sets; P demand_predl , P diesel , P motor , P battery They are the predicted demand power, ship diesel engine reference power, motor reference power and battery reference power.

[0063] Step 6: The droop controller dynamically updates the droop coefficient in real time according to step 5. The droop controller calculates the reference voltage of different power sources according to the output current of different power sources.

[0064] V ref,generator1 (t) = V0 - k generator1 I generator1 (t)

[0065] V ref,generator2 (t) = V0 - k generator2 I generator2 (t)

[0066] V ref,battery(t) = V0 - k battery I battery (t)

[0067] Step 7: The power control layer obtains the average voltage V of the distributed power source based on the consistency convergence matrix according to the deviation between the output voltage value of different power sources and the expected value of the bus voltage. ave , the compensation voltage values ​​of different power sources are obtained according to the PI controller.

[0068] ΔU generator1 =k p1 (V0-V ave )+k i1 ∫(V0-V ave )dt

[0069] ΔU generator2 =k p2 (V0-V ave )+k i2 ∫(V0-V ave )dt

[0070] ΔU battery =k p3 (V0-V ave )+k i3 ∫(V0-V ave )dt

[0071] Step 8: The power source reference voltage obtained in step 6 and the power source compensation voltage obtained in step 7 are superimposed to obtain a final reference control voltage of the power source.

[0072] V ref_new,generator1 (t) = V ref,generator1 (t)+ΔU generator1

[0073] V ref_new,generator2 (t) = V ref,generator2 (t)+ΔU generator2

[0074] V ref_new,battery (t) = V ref,battery (t)+ΔU battery

[0075] Step 9: According to step 8, the final reference control voltage of the power source is obtained and transmitted to the corresponding rectifier and converter control units respectively, and the rectifier and converter are controlled to output the corresponding voltage, thus completing one control.

[0076] Step 10: Repeat steps 1 to 9 to implement the above control process in real-time closed loop, and optimize the energy consumption of the ship hybrid power system in real time based on the proposed hierarchical distributed control method.

[0077] The ship diesel engine and the ship motor are mechanically connected in parallel via a gearbox; the generator set and the battery are electrically connected in parallel to form a ship power grid bus, which is used to supply energy to the ship motor.

[0078] The hierarchical distributed control method comprises three layers: an energy management layer, a power control layer, and a local control layer. The energy management layer includes a real-time optimized energy management strategy combining forward dynamic solution and model predictive control (FDP-MPC); the power control layer includes a distributed control method based on a consensus algorithm; and the local control layer includes a droop control method based on dynamic virtual impedance.

[0079] It will be easily understood by those skilled in the art that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A hierarchical distributed control method for a multi-energy ship hybrid power system, characterized in that: The following steps are involved: S1. According to the ship load prediction model, predict k +1 to k +N ship power requirement within time; S2, using the forward dynamic programming algorithm, with the minimum equivalent fuel consumption as the optimization goal, calculate k +1 to k +N time period: the reference charge and discharge power of the battery, the reference output power of the diesel engine, the reference output power of the motor, and the reference output power of the generator set; S3. According to k +1 to k The reference charge and discharge power of the battery within +N time is calculated k +1 to k + The battery SOC trajectory within N moments; S4, using the model predictive control method to track the SOC trajectory through the model predictive controller. The model predictive control method is based on the established hybrid system state equation to track the SOC trajectory, with the minimum tracking error as the optimization goal, and calculates k +1 to k +N time period: the new battery reference charge and discharge power, diesel engine reference output power, motor reference output power, and generator set reference output power; S5. According to the deviation between the output voltage of the power source and the expected voltage, the reference control voltage compensation amount Δ of each power source is calculated based on the consistency algorithm. U ; S6. Dynamically calculate the droop coefficient based on the new reference charge and discharge power of the battery, the reference output power of the diesel engine, the reference output power of the motor, and the reference output power of the generator set. The real-time droop coefficient calculation formula is: in, k i For the i The droop coefficient of each power source; p motor is the new motor reference output power; p battery Is the reference charge and discharge power of a new battery. p battery >0, discharge, when p battery <0, then charge; It is i The new reference power of each power source; The reference control voltage of each power source is calculated by the droop controller V ref The droop controller calculates the reference control voltage of each power source through the droop control method based on dynamic virtual impedance. V ref ; S7, reference control voltage V ref and reference control voltage compensation Δ U Perform superposition processing as the final control voltage output to control the output voltage of each power source and realize hierarchical distributed control.

2. A hierarchical distributed control method for a multi-energy ship hybrid power system according to claim 1, characterized in that: The hybrid system state equation is established based on the relationship between the battery's SOC trajectory and the charge and discharge power and time; the SOC trajectory is the relationship between the battery's remaining power and time.

3. The hierarchical distributed control method for a multi-energy ship hybrid power system according to claim 1 is characterized in that: The forward dynamic programming algorithm in S2 is based on the current system SOC state and the predicted ship power demand, and is discretized according to the characteristics of the battery. k +1 to k The possible output power of the battery within +N time is traversed and optimized to find the optimal solution; the optimization goal of the optimization process is to minimize the equivalent fuel consumption.

4. The hierarchical distributed control method for a multi-energy ship hybrid power system according to claim 1, characterized in that: The load prediction in the ship load prediction model in S1 is implemented based on a machine learning intelligent algorithm; its prediction time domain is set according to the accuracy of the ship operating condition prediction.

5. The hierarchical distributed control method for a multi-energy ship hybrid power system according to claim 1, characterized in that: S5 is based on the deviation between the ship bus voltage and the expected value. By adjusting and compensating the voltage output by the local controller, the ship grid bus voltage is brought close to the expected value. The consistency algorithm communicates and exchanges information based on the output voltage of each power source to obtain the reference control voltage compensation. The reference control voltage compensation is obtained through PI control and its calculation method is: in, is the reference control voltage compensation output by the distributed controller; It is i The output voltage of each power source; and These are PI controller parameters; is the average voltage of each power source obtained based on the consistency algorithm; is the expected value of the ship power grid bus voltage.

6. The hierarchical distributed control method for a multi-energy ship hybrid power system according to claim 5 is characterized in that: The calculation method is: in, For the discretized k Output voltage matrix of distributed power sources at all times; D It is the consistency matrix determined according to the topology of the controlled object based on consistency theory.

7. The hierarchical distributed control method for a multi-energy ship hybrid power system according to claim 5, characterized in that: The principle of droop control based on dynamic virtual impedance droop control method is: in, for t Moment, i A reference control voltage of a power source; k i For the i The droop coefficient of a power source.

8. A system using the hierarchical distributed control method for a multi-energy ship hybrid power system according to any one of claims 1 to 7, characterized in that: It includes the energy management layer, power control layer and local control layer which are arranged in sequence; Energy management layer, used to predict the ship load based on the ship load prediction model k +1 to k +N time required power of the ship; using the forward dynamic programming algorithm, with the minimum equivalent fuel consumption as the optimization goal, calculate k +1 to k +N time within the battery reference charge and discharge power, diesel engine reference output power, motor reference output power and generator reference output power; according to k +1 to k The reference charge and discharge power of the battery within +N time is calculated k +1 to k The SOC trajectory of the battery within +N moments; the SOC trajectory is tracked by the model predictive controller, with the minimum tracking error as the optimization goal, and the calculation is obtained k +1 to k +N time period: the new battery reference charge and discharge power, diesel engine reference output power, motor reference output power, and generator set reference output power; The power control layer is used to calculate the reference control voltage compensation value Δ of each power source based on the deviation between the output voltage of the power source and the expected voltage using a consistency algorithm. U ; The local control layer is used to dynamically calculate the droop coefficient based on the new reference charge and discharge power of the battery, the reference output power of the diesel engine, the reference output power of the motor, and the reference output power of the generator set; the reference control voltage of each power source is calculated by the droop controller V ref ; For reference control voltage V ref and reference control voltage compensation Δ U Perform superposition processing as the final control voltage output to control the output voltage of each power source and realize hierarchical distributed control.

Images