Multi-mode control and reactive voltage on-site management and control method for port hybrid energy supply system

Abstract

The invention relates to a multi-mode control and reactive voltage local management and control method for a port hybrid energy supply system, and belongs to the technical field of application of voltage control in an electric power system. The method comprises the following steps: S1, analyzing a power flow relation of a direct current micro-grid under different operation conditions, and dividing a system into five working modes according to a bus voltage fluctuation range; s2, setting a voltage recovery control module and a current adjustment control module; s3, according to photovoltaic output and load day-ahead prediction of historical statistical data, constructing a probability model of photovoltaic output and load in each time period to determine a day-ahead adjustment plan of on-load voltage regulating transformer tapping, capacitor bank and photovoltaic converter reactive output; and S4, adjusting the reactive power adjustment amount of the inverter and the active power of the photovoltaic inverter. According to the method, the switching flexibility of multiple working modes of the port micro-grid is improved, the voltage fluctuation out-of-limit and the network loss increase are reduced, and the economical efficiency of the port power distribution network is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of application of voltage control in power systems, and relates to a multi-mode control method of a port hybrid energy supply system and a method for local control of reactive voltage. Background Art

[0002] With the rapid development of inland green ports, ports, as the assembly point and hub of shipping and land transportation, play a key role in the development of the national economy. The scale of green energy power generation in ports has also increased year by year. Among them, photovoltaic power generation is particularly important in inland ports. By combining energy storage, it plays an important role in reducing the operating cost of port microgrids and reducing carbon emissions. How to consider the online transfer of current and the seamless switching of operation modes under the complex operation of port microgrids has attracted the attention of domestic and foreign scholars. At the same time, the increase in network losses and voltage over-limit caused by the addition of large-scale distributed power sources to the power grid, and the problem that traditional voltage regulating devices cannot frequently respond to rapid voltage changes and the poor effect of smoothing voltage fluctuations also need to be solved. Optimizing the unified control system architecture of the flexible switching of various working modes of the port microgrid, considering the multi-source load local control strategy of new energy consumption and multiple charging load priority sequences can effectively improve the operation stability and new energy utilization rate of the port microgrid. Summary of the invention

[0003] The present invention provides a multi-modal control and reactive voltage on-site management and control method for a port hybrid energy supply system, which improves the flexibility of switching between multiple working modes of the port microgrid, reduces voltage fluctuation over-limit and network loss increase, and improves the economy of the port distribution network.

[0004] A multi-mode control and reactive voltage local control method for a port hybrid energy supply system comprises the following steps:

[0005] Step S1, analyzing the power flow relationship of the DC microgrid under different operating conditions, and dividing the system into five working modes according to the bus voltage fluctuation range;

[0006] The overall coordinated control function of each unit in the system is realized through step S1. The following section will give in detail the control strategy of each unit under each operating condition;

[0007] Step S2, setting two modules: voltage recovery control and current adjustment control;

[0008] Step S3: Based on the day-ahead forecast of photovoltaic output and load from historical statistical data, a probability model of photovoltaic output and load in each time period is constructed to determine the day-ahead adjustment plan for the reactive output of the on-load tap changer (OLTC), capacitor bank and photovoltaic transformer;

[0009] Step S4: utilizing the dynamic reactive power response capability of the photovoltaic inverter to adjust the reactive power adjustment amount of the inverter and the active power of the photovoltaic inverter to suppress voltage over-limit and fluctuation.

[0010] Optionally, step S1 specifically includes:

[0011] Step S11, analyzing the power flow relationship of the port DC microgrid, including the power relationship under various operating conditions;

[0012] Step S12: Based on the DC bus voltage control, different working modes are divided according to the power flow relationship of the port microgrid under different working conditions to achieve distributed control of energy storage, photovoltaic system and grid-connected converter, so as to achieve stable system operation.

[0013] Optionally, step S11 specifically includes:

[0014] S111. The power balance relationship of the DC microgrid system under the grid-connected operation condition of the port DC microgrid is: P S =P PV -P bes -P load , where P S Grid exchange power, P PV is the photovoltaic power generation power, P bes is the energy storage exchange power, P load is the load power;

[0015] S112, AC grid-connected converter (GCC) works in rectification and inversion mode. If the AC system and PV unit cannot meet the power requirements of the load even at full power or maximum power output, the energy storage system will provide the remaining required power and stabilize the DC bus voltage. At this time, the system enters a new stable operation state, and the power provided by the energy storage system P Bes For: P Bes =P Smax +P PV -P load , where P Smax is the maximum value of the power exchanged by the power grid;

[0016] S113. If the load cannot completely consume the power generated by the distributed power source, the grid-connected converter enters the inverter working state. When the power flowing through the grid-connected converter reaches the limit value, and the sunlight is very strong or the remaining power of the energy storage system is not full, the photovoltaic and energy storage units need to jointly take over the grid-connected converter as the system's relaxation terminal to maintain the DC bus voltage. The power balance relationship at this time is: P Smax =P PV -P load -P Bes ;

[0017] S114. When the port DC microgrid is in isolated operation, the energy storage system is the only balance point of the system, and the power relationship is: P Bes =P PV -P load .

[0018] Optionally, the step S12 is specifically:

[0019] Working mode 1: U dc ≤0.95U ref The energy storage unit is responsible for maintaining the bus voltage stable and discharging to the system. The distributed power source works in MPPT mode. dc is the DC bus voltage, U ref is the rated voltage;

[0020] Working mode 2: 0.95U ref ≤U dc ≤0.98U ref , when the photovoltaic power generation is in the MPPT state and is not enough to provide the power required by the load, the DC bus voltage is lower than the rated voltage, and the AC grid supplements the power shortage to the DC microgrid through the grid-connected converter at the maximum power. The energy storage unit acts as a relaxation terminal and discharges. When the SOC drops to the minimum limit, the discharge should be stopped;

[0021] Working mode 3: 0.98U ref ≤U dc ≤1.02U ref At this time, the grid-connected converter and energy storage unit adopt voltage droop control as relaxation terminals, the distributed power source adopts MPPT control, and the load power is provided by the distributed power source;

[0022] Working mode 4: 1.02U ref ≤U dc ≤1.05U ref At this time, the DC bus voltage is higher than the rated voltage, the grid-connected converter switches to the current limiting control state, the photovoltaic unit still operates in MPPT mode, and the energy storage unit acts as a relaxation terminal to control the DC bus voltage stability;

[0023] Working mode 5: U dc ≥1.05U ref , there is too much residual power in the system, the SOC of the energy storage battery reaches the upper limit, and the photovoltaic unit adopts droop control.

[0024] Optionally, in step S12, distributed control is performed on the energy storage, photovoltaic system and grid-connected converter, and step S12 is specifically as follows:

[0025] Step S121: Set the droop control to be used in grid-connected and islanded operation to maintain the DC bus voltage stable. When the AC main grid fails, the microgrid needs to operate in an islanded operation for a short time. At this time, the energy storage unit will serve as a balancing node to maintain the system power balance and the stability of the DC bus voltage.

[0026] Step S122, photovoltaic power generation unit control strategy, the output voltage of the photovoltaic power generation system is lower than the DC bus voltage, and the photovoltaic interface converter is connected to the DC microgrid via a BOOST type DC / DC converter;

[0027] Step S123, the AC grid-connected coverter (GCC) is a voltage type converter, the AC grid-connected coverter is regulated by a PI regulator, the AC grid-connected coverter is used to provide a DC bus voltage and maintain a DC grid power balance, the AC grid-connected coverter adopts direct current control, the outer loop is a voltage loop, which is used to control the DC bus voltage stability; the inner loop is used to control the current.

[0028] Optionally, the step S2 is specifically:

[0029] Step S21: adding distributed collaborative secondary control including voltage recovery control and current correction control to the DC microgrid control system based on discrete consistency theory;

[0030] Step S22: adding a consistency gain function to reduce noise.

[0031] Optionally, step S21 is specifically:

[0032] Step S211, the distributed collaborative secondary control is a double closed-loop control, the double closed-loop control is a current inner loop control and a voltage outer loop return voltage droop control, the local controllers of each microgrid establish a communication connection, each controller only collects its own voltage and output current information, and only exchanges information with the controllers of adjacent microgrids;

[0033] Step S212, obtaining the average voltage and average current value of each node in the entire system through consistency iteration of the information of the dual closed-loop control;

[0034] Step S213, the current regulator and voltage regulator subtract the average current value and voltage value output by the consistency algorithm from the actual current value and voltage value of each node, and then amplify them through a PI controller to obtain virtual current and voltage increments to compensate for the bus voltage drop of the entire system.

[0035] Optionally, the consistency gain function is:

[0036]

[0037] Where j is the state variable input interfered by random communication noise, and the random communication noise N j and J i It can be represented by a Gaussian white noise, which is a uniformly bounded independent random variable with a mean of 0. ij is the gain coefficient, x i (k), x j (k), v i (k) are the noises in the sending, transmission and receiving stages at time k, respectively, and x i (k+1) is the noise in the sending stage at time k+1, the consistency gain function a(k), and a(k) takes the value of ±1, ±2 or ±3.

[0038] Optionally, step S3 specifically includes:

[0039] Step S31, reactive power optimization takes the grid voltage over-limit risk, network loss and voltage deviation as the minimum objective function, min F = λ1f1 + λ2f2 + λ3f3, where f1 is the cumulative membership function of all node voltage over-limit risks at each moment, f2 is the membership function of the expected network loss index, and f3 is the membership function of the voltage deviation index;

[0040]

[0041] Among them, S i,t is the voltage over-limit indicator value, V min 、V max and V B They are the maximum allowable voltage, the minimum allowable voltage and the reference voltage respectively; f i,t (x) is the voltage probability distribution of the node at time, t and T are time, i and N are nodes, x and dx are voltage and voltage differential variables;

[0042]

[0043] Among them, N b is the number of branches; P i,t and Q i,t are the active power and reactive power of branch i at time t respectively; R i and U i,t are the resistance of branch i and the voltage at the first end at time t respectively;

[0044]

[0045] Among them, V i N is the rated voltage of node i; V i,t is the voltage of node i at the moment;

[0046] Fuzzy set theory is used to describe the optimization result of the objective function using the membership function F, whose value range is between 0 and 1. The smaller F is, the closer it is to the optimal value, indicating that the optimization result is more ideal. The weights λ1, λ2 and λ3 are determined using the hierarchical analysis method.

[0047] Step S32, constructing constraint conditions, including: power balance constraint, energy storage processing constraint, transformer constraint, photovoltaic output constraint and photovoltaic inverter reactive output constraint;

[0048] Step S33, read multiple scenario sets, each scenario set includes load and new energy output, the scenario set can be a microgrid, combine the objective function F and the constraint conditions, and output the day-ahead regulation plan of the on-load tap-changing transformer tap, capacitor bank and photovoltaic transformer reactive output;

[0049] Step S4 is specifically as follows:

[0050] Step S41, adjusting the active power output increment of the photovoltaic inverter according to the photovoltaic active power output prediction error;

[0051] Step S42, superimposing the active power output increment of the photovoltaic inverter with the day-ahead planned value of the period to obtain the reactive power output of the inverter;

[0052] Step S43: When it is detected that the voltage at the photovoltaic grid point exceeds the limit, Calculate the inverter reactive power adjustment required for PV inverter voltage regulation in, is the voltage reactive sensitivity of node i in period t, V i,τ is the grid connection point voltage value, is the voltage limit of node i.

[0053] According to one aspect of the present application, a multi-modal control and reactive voltage local control method control system of a port hybrid energy supply system includes:

[0054] at least one processor; and,

[0055] a memory communicatively connected to at least one of the processors; wherein,

[0056] The memory stores instructions that can be executed by the processor, and the instructions are used to be executed by the processor to implement the multi-modal control and reactive voltage on-site control method of the port hybrid energy supply system as described in any of the above items.

[0057] Beneficial effects:

[0058] Through the multi-modal control and reactive voltage on-site control method of the port hybrid energy supply system of the present invention, the flexibility of switching between various working modes of the port microgrid is improved, the voltage fluctuation limit and network loss increase are reduced, and the economy, stability and new energy utilization rate of the port distribution network are improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 This is a power flow relationship diagram of a port microgrid according to an embodiment of the present invention.

[0060] Figure 2 It is a schematic diagram of the flow chart of the improved hybrid frog leaping algorithm (SFLA) according to an embodiment of the present invention.

[0061] Figure 3 It is a schematic diagram of the process of real-time on-site control of reactive voltage in a port microgrid according to an embodiment of the present invention. DETAILED DESCRIPTION

[0062] The present invention is further described below in conjunction with the accompanying drawings and specific embodiments.

[0063] Embodiment 1

[0064] A multi-mode control and reactive voltage local control method for a port hybrid energy supply system comprises the following steps:

[0065] Step S1, analyzing the power flow relationship of the DC microgrid under different operating conditions, and dividing the system into five working modes according to the bus voltage fluctuation range;

[0066] The overall coordinated control function of each unit in the system is realized through step S1. The following section will give in detail the control strategy of each unit under each operating condition;

[0067] Step S2, setting two modules: voltage recovery control and current adjustment control;

[0068] Step S3: Based on the day-ahead forecast of photovoltaic output and load from historical statistical data, a probability model of photovoltaic output and load in each time period is constructed to determine the day-ahead adjustment plan for the reactive output of the on-load tap changer (OLTC), capacitor bank and photovoltaic transformer;

[0069] Step S4: utilizing the dynamic reactive power response capability of the photovoltaic inverter to adjust the reactive power adjustment amount of the inverter and the active power of the photovoltaic inverter to suppress voltage over-limit and fluctuation.

[0070] Preferably, as described in step S1, the power flow relationship of the port DC microgrid under different operating conditions is analyzed, and then the system is divided into five working modes according to the bus voltage fluctuation range, so as to realize the overall coordinated control function of each unit in the system, and give the control strategy of each unit under each operating condition in detail. Step S1 is specifically as follows:

[0071] Step S11, analyzing the power flow relationship of the port DC microgrid, including the power relationship under various operating conditions;

[0072] Step S12: Based on the DC bus voltage control, different working modes are divided according to the power flow relationship of the port microgrid under different working conditions to realize the distributed control of energy storage, photovoltaic system and grid-connected converter, so as to achieve the stability of system operation;

[0073] Preferably, the step S11 is specifically as follows:

[0074] S111. The power balance relationship of the DC microgrid system under the grid-connected operation condition of the port DC microgrid is: P S =P PV -P bes -P load , where P S Grid exchange power, P PV is the photovoltaic power generation power, P bes is the energy storage exchange power, P load is the load power;

[0075] S112, AC grid-connected converter (GCC) works in rectification and inversion mode. If the AC system and PV unit cannot meet the power requirements of the load even at full power or maximum power output, the energy storage system will provide the remaining required power and stabilize the DC bus voltage. At this time, the system enters a new stable operation state, and the power provided by the energy storage system P Bes For: P Bes =P Smax +P PV -P load , where P Smax is the maximum value of the power exchanged by the power grid;

[0076] S113. If the load cannot completely consume the power generated by the distributed power source, the grid-connected converter enters the inverter working state. When the power flowing through the grid-connected converter reaches the limit value, and the sunlight is very strong or the remaining power of the energy storage system is not full, the photovoltaic and energy storage units need to jointly take over the grid-connected converter as the system's relaxation terminal to maintain the DC bus voltage. The power balance relationship at this time is: P Smax =P PV -P load -P Bes ;

[0077] S114. When the port DC microgrid is in isolated operation, the energy storage system is the only balance point of the system, and the power relationship is: P Bes =P PV -P load .

[0078] Optionally, in step S12, the port microgrid operating mode is divided, and the step S12 is specifically:

[0079] Working mode 1: U dc ≤0.95U ref The energy storage unit is responsible for maintaining the bus voltage stable and discharging to the system. The distributed power source works in MPPT mode. dc is the DC bus voltage, U ref is the rated voltage;

[0080] Working mode 2: 0.95U ref ≤U dc ≤0.98U ref , when the photovoltaic power generation is in the MPPT state and is not enough to provide the power required by the load, the DC bus voltage is lower than the rated voltage, and the AC grid supplements the power shortage to the DC microgrid through the grid-connected converter at the maximum power. The energy storage unit acts as a relaxation terminal and discharges. When the SOC drops to the minimum limit, the discharge should be stopped;

[0081] Working mode 3: 0.98U ref ≤U dc ≤1.02U ref At this time, the grid-connected converter and energy storage unit adopt voltage droop control as relaxation terminals, the distributed power source adopts MPPT control, and the load power is provided by the distributed power source;

[0082] Working mode 4: 1.02U ref ≤U dc ≤1.05U ref At this time, the DC bus voltage is higher than the rated voltage, the grid-connected converter switches to the current limiting control state, the photovoltaic unit still operates in MPPT mode, and the energy storage unit acts as a relaxation terminal to control the DC bus voltage stability;

[0083] Working mode 5: U dc ≥1.05U ref , there is too much residual power in the system, the SOC of the energy storage battery reaches the upper limit, and the photovoltaic unit adopts droop control.

[0084] The working mode is divided based on the DC bus voltage control in order to achieve the following goals: make full use of the electric energy generated by renewable energy, and cooperate with the energy storage unit to realize the coordinated control and optimal energy configuration among the micro sources of the port under different voltage levels; realize the distributed control of the system to ensure the stability of the system operation. Therefore, in each operating mode, at least two relaxation terminals are set to jointly control the DC bus voltage stability according to the voltage droop characteristics to prevent the system from being instability due to a fault at one end. The voltage reference critical value is set to not exceed the maximum voltage deviation range allowed by the DC bus.

[0085] Optionally, in step S12, distributed control is performed on the energy storage, photovoltaic system and grid-connected converter, and step S12 is specifically as follows:

[0086] Step S121, the energy storage unit control strategy is based on the characteristics that the output of the photovoltaic unit is unstable, while the output power of the battery is relatively stable. It is set to use droop control in grid-connected and islanded operation to maintain the stability of the DC bus voltage. When the AC main grid fails, the microgrid needs to operate in an islanded operation for a short time. At this time, the energy storage unit will serve as a balancing node to maintain the system power balance and the stability of the DC bus voltage;

[0087] In order to avoid overcharge and overdischarge of the energy storage unit and improve the service life of the battery, a capacity over-limit protection strategy is added to limit the rated range of SOC to 40%-90%;

[0088] Step S122, photovoltaic power generation unit control strategy, the output voltage of the photovoltaic power generation system is lower than the DC bus voltage, and the photovoltaic interface converter is connected to the DC microgrid via a BOOST type DC / DC converter.

[0089] There are two control modes for photovoltaic interface converters. In order to make full use of distributed energy, PV units usually use maximum power tracking (MPPT) control. In this example, the perturbation observation method is used to implement the MPPT of the photovoltaic unit. The judgment steps are as follows:

[0090] a) Measure the photovoltaic output voltage and current and calculate the output power;

[0091] b) Compare the output power with the output power at the previous moment, and adjust the operating point based on the comparison result.

[0092] c) If the output power of the photovoltaic cell is greater than the previous value, the voltage is compared. If the voltage is greater, the voltage is made greater. If the voltage is smaller, the voltage is made smaller.

[0093] d) If it is less than the previous power, it means that the control will reduce the output power, so the voltage will change in the opposite direction;

[0094] e) Return to step a) and repeat the process.

[0095] In some special cases, such as when a PCC point fails, in order to reduce the impact of the failure on the system, the DC microgrid needs to be in an islanded operation state for a short time, and the grid-connected converter cannot control the DC bus voltage to be constant. At this time, multiple distributed power sources need to jointly stabilize the DC bus voltage. In addition, if the output power of the PV unit is much greater than the load consumption in the grid-connected operation state, there will be excess power in the DC microgrid. In order to avoid excessive DC bus voltage in the above situation, this example sets the second control mode of the photovoltaic system, constant voltage droop control;

[0096] Step S123, the AC grid-connected coverter (GCC) is a voltage type converter, the AC grid-connected coverter is regulated by a PI regulator, the AC grid-connected coverter is used to provide a DC bus voltage and maintain a DC grid power balance, the AC grid-connected coverter adopts direct current control, the outer loop is a voltage loop, which is used to control the DC bus voltage stability; the inner loop is used to control the current.

[0097] In summary, in step S1, the function of overall coordinated control of each unit in the system is realized, the system is divided into five working modes according to the bus voltage fluctuation range, and the control strategy of each unit under each operating condition is given in detail, making it clear that each unit can meet the normal operation of the port microgrid under different working conditions, and can enable the system to switch smoothly between different operating modes, and can maintain real-time power balance and voltage stability.

[0098] According to one aspect of the present application, the multimodal control and reactive voltage local control method of the port hybrid energy supply system as described in step S2 draws on the discrete consistency theory used for sensor networks and designs two modules of voltage recovery control and current adjustment control; secondly, by adding a consistency gain function, the influence on the secondary optimization control effect under non-ideal sparse communication conditions is effectively reduced, and the step S2 is specifically as follows:

[0099] Step S21: adding distributed collaborative secondary control including voltage recovery control and current correction control to the DC microgrid control system based on discrete consistency theory;

[0100] Step S22, adding a consistency gain function to reduce noise;

[0101] Preferably, step S21 is specifically:

[0102] Step S211, the distributed collaborative secondary control is a double closed-loop control, the double closed-loop control is a current inner loop control and a voltage outer loop return voltage droop control, the local controllers of each microgrid establish a communication connection, each controller only collects its own voltage and output current information, and only exchanges information with the controllers of adjacent microgrids;

[0103] Step S212, obtaining the average voltage and average current value of each node in the entire system through consistency iteration of the information of the dual closed-loop control;

[0104] Step S213: The current regulator and the voltage regulator make a difference between the average current value and voltage value output by the consistency algorithm and the actual current value and voltage value of each node, and then amplify them through a PI controller to obtain virtual current and voltage increments to compensate for the bus voltage drop of the whole system;

[0105] According to one aspect of the present application, the discrete consensus algorithm used in step S212 is based on the local information transmission between each agent and its neighboring nodes, so that the individuals finally reach a consensus on a state variable according to the distributed control rules, which is specifically:

[0106] In this example where there are many distributed power sources, the distributed power sources will be affected by communication noise during the sending, transmission, and receiving stages of the measurement signal, which will reduce the convergence of the consistency algorithm. Therefore, this article adopts the following consistency algorithm:

[0107]

[0108] Where j is the state variable input interfered by random communication noise, and the random communication noise N j and J i It can be represented by a Gaussian white noise, which is a uniformly bounded independent random variable with a mean of 0. ij is the gain coefficient, x i (k), x j (k), v i (k) are the noises in the sending, transmission and receiving stages at time k, respectively, and x i (k+1) is the noise in the sending phase at time k+1.

[0109] When the convergence coefficient ε is a constant, it is not robust to Gaussian noise. Therefore, in a non-ideal communication environment, in order to ensure that the consistency error converges to 0, this paper adopts an attenuated consistency gain function a(k), and the value of a(k) is ±1, ±2 or ±3.

[0110]

[0111] In step S2, a distributed collaborative secondary control including voltage recovery control and current correction control is added to the DC microgrid control system based on discrete consistency theory, and a consistency gain function is introduced to eliminate the influence of noise. This method solves the serious defect that in the primary control of the traditional DC microgrid system, multiple distributed power sources with droop characteristics are usually selected to share the load power. This control method with a fixed droop coefficient cannot simultaneously meet the two control objectives of proportional current distribution and bus voltage stability. The main reasons for this defect are the following two points: 1) The line impedance difference is too large, which will cause the load to be unable to be reasonably distributed; 2) The droop control is a differential adjustment. The addition of virtual impedance increases the voltage deviation of each distributed unit node, which is easy to exceed the allowable offset range.

[0112] The quadratic optimization control method can quickly track and respond to keep the current output consistent in the case of unknown line impedance, load changes, and distributed power fluctuations under ideal communication environment, and has strong adaptability. When the sparse communication network is disturbed by noise, the influence of noise on the consistency convergence accuracy is effectively suppressed by adding a consistency gain function, which improves the reliability of the quadratic optimization control.

[0113] Optionally, step S3 specifically includes:

[0114] Step S31, reactive power optimization takes the grid voltage over-limit risk, network loss and voltage deviation minimization as the objective function, min F = λ1f1+λ2f2+λ3f3, where f1 is the cumulative membership function of the voltage over-limit risk of all nodes at each moment, f2 is the membership function of the expected network loss index, and f3 is the membership function of the voltage deviation index.

[0115]

[0116] Among them, S i,t is the voltage over-limit indicator value, V min 、V max and V B They are the maximum allowable voltage, minimum allowable voltage and reference voltage respectively. i,t (x) is the voltage probability distribution of the node at time, t and T are time, i and N are nodes, x and dx are voltage and voltage differential variables;

[0117]

[0118] Among them, N b is the number of branches; P i,t and Q i,t are the active power and reactive power of branch i at time t respectively; R i and U i,t are the resistance of branch i and the voltage at the first end at time t respectively.

[0119]

[0120] Among them, V i N is the rated voltage of node i; V i,t is the voltage at node i at the moment.

[0121] Fuzzy set theory is used to describe the optimization result of the objective function using the membership function F, whose value range is between 0 and 1. The smaller F is, the closer it is to the optimal value, indicating that the optimization result is more ideal. The weights λ1, λ2 and λ3 are determined using the hierarchical analysis method.

[0122] Optionally, λ1+λ2+λ3=1, and it can be understood that min is the minimum value.

[0123] Step S32: constructing constraint conditions, including: power balance constraint, energy storage processing constraint, transformer constraint, photovoltaic output constraint and photovoltaic inverter reactive output constraint.

[0124] It can be understood that the constraints include power limit, energy storage battery SOC value, transformer power limit, photovoltaic output limit and photovoltaic inverter reactive output limit.

[0125] Step S33, read multiple scenario sets, each scenario set includes load and new energy output, the scenario set can be a microgrid, combine the objective function F and the constraint conditions, and output the day-ahead regulation plan of the on-load tap changer (OLTC), capacitor bank and photovoltaic inverter reactive output;

[0126] Optionally, step S33 is specifically as follows: step S34, relaxing the discrete variables into continuous variables, using the improved frog leaping algorithm to perform 24h static optimization to obtain the continuous values ​​of the 24h optimal adjustment plan of each capacitor bank and the on-load tap changer OLTC, completing the time series segmentation of the capacity curve of each capacitor bank and the gear curve of the on-load tap changer OLTC tap by Word system clustering, and calculating the capacity or gear in each segment;

[0127] Step S35, fix the discrete device adjustment plan, use the improved hybrid frog leaping (SFLA) algorithm to solve the reactive power output of the photovoltaic inverter, and finally complete the reactive power dispatch plan on the day before;

[0128] Step S36, result analysis and verification, including: using a cross-validation method to evaluate the generalization ability of the model and obtain a verification index V;

[0129] Conduct sensitivity analysis to evaluate the impact of key parameters on the results and obtain the sensitivity matrix S;

[0130] The Monte Carlo simulation method is used to generate random scenarios for model robustness testing and obtain the robustness index R;

[0131] Output verification index V, sensitivity matrix S and robustness index R.

[0132] Preferably, step S35 is specifically:

[0133] Step S351, performing Word system clustering on the optimal gear position of the on-load tap-changing transformer and the reactive output sequence of the capacitor bank within 24 hours obtained in the previous step;

[0134] Step S352, determining the final solution of the 24h on-load tap-changing transformer gear position and the reactive output of the capacitor bank;

[0135] Step S353, fixing the gear position of the on-load tap-changing transformer and the reactive output of the capacitor bank, and solving the reactive output of the photovoltaic inverter using a particle swarm;

[0136] Step S354, outputting the reactive power / voltage optimization control result of the day before;

[0137] Optionally, step S353 is specifically:

[0138] Step S3531: firstly, arrange the 24-hour action gears of the discrete devices in chronological order and regard them as a set of ordered sample sets.

[0139] Step S3532: Calculate the sum of squares of deviations after merging two adjacent samples. Where W is the sum of squares of deviations; n d H is the number of new samples formed by merging adjacent samples; d is the dth sample in the new sample; H^ is the average value of the new sample, for The transpose of

[0140] Step S3533: select two samples with the smallest sum of squared deviations to merge; if there are two or more adjacent samples with the same and smallest sum of squared deviations, merge them into one class.

[0141] Step S3534: The multiple samples after merging are regarded as a new sample, and the order of the time periods is kept unchanged to update the sample set.

[0142] Step S3535, return to step S3532 until the number of clusters is the same as the number of discrete device action constraints.

[0143] Step S3536: regularize the average values ​​of the action gears in each category to obtain the adjustment gear of the discrete device.

[0144] Optionally, the improved hybrid frog leaping algorithm (SFLA) used in step S35 is specifically:

[0145] Step S321, randomly generate M frog individuals, which constitute the initial population P = {X1, X2, ..., XM}, the i-th frog in the s-dimensional solution set is represented by Xi = (xi1, xi2, ..., xis), set the local search number and the global maximum convergence number;

[0146] Step S322, first calculate the fitness of all individuals, and arrange them in descending order, and record the minimum fitness value as Xglo;

[0147] Step S323, then divide the population into small sub-populations according to the rules, and divide them into m sub-populations on average. Each sub-population contains n individuals, and their relationship is M=m×n. The allocation rule is: the first frog enters the sub-population group No. 1, the second frog enters the sub-population group No. 2, and the allocation continues. The mth frog is divided into the mth sub-population group, and the m+1th frog is re-divided into the 1st sub-population group, and so on, and the allocation is average. In this way, the minimum fitness value in each sub-population is recorded as Xb, and the worst fitness is recorded as Xw;

[0148] Step S324, updating the position according to the divided sub-species groups according to the update strategy, wherein the position is improved by introducing an intermediate factor and an acceleration factor;

[0149] a) Take the average value of all individuals except the worst frog as the intermediate factor Xa, then remix them and divide them into two populations mX1 and mX2, record Xavel as the average value of population 1, and Xave2 as the average value of population 2;

[0150] b) If the fitness value of Xavel is worse than the fitness value corresponding to the intermediate factor when compared with the intermediate factor, the population is eliminated, and the uneliminated population is selected to update the worst frog, and then the above operation is repeated in the uneliminated population;

[0151] c) The formula used for location update is: St=C·rand()×(Xb-Xave), Xnew=Xw+St(Stmin≤St≤Stmax), where St is the worst frog change step length, C is the acceleration factor, rand() function represents the generation of uniformly distributed random real numbers between 0 and 1, Xi is the other individuals except Xw, Stmin and Stmax are the minimum and maximum step lengths respectively;

[0152] Step S325: After the algorithm is updated by the above formula, if the frog Xnew is better than the original frog Xw, then replace the frog in the original sub-population group; if the performance has not improved, replace Xb with Xglo, and the algorithm updates the position of the worst frog again; if after this update, the fitness value is still worse than the value before the update, then randomly select any frog individual in the population to replace the original Xw.

[0153] Step S326, when the number of local searches reaches the upper limit allowed, remix the frogs in all sub-species groups and repeat the above optimization process;

[0154] Step S327, repeat this process until the global maximum number of convergences is reached;

[0155] Optionally, step S4 specifically includes:

[0156] Step S41, adjusting the active power output increment of the photovoltaic inverter according to the photovoltaic active power output prediction error;

[0157] Step S42, superimposing the active power output increment of the photovoltaic inverter with the day-ahead planned value of the period to obtain the reactive power output of the inverter;

[0158] Step S43: When it is detected that the voltage at the photovoltaic grid point exceeds the limit, Calculate the inverter reactive power adjustment required for PV inverter voltage regulation in, is the voltage reactive sensitivity of node i in period t, V i,τ is the grid connection point voltage value, is the voltage limit of node i.

[0159] According to one aspect of the present application, a multi-modal control and reactive voltage local control method control system for a port hybrid energy supply system is characterized by comprising:

[0160] at least one processor; and,

[0161] a memory communicatively connected to at least one of the processors; wherein,

[0162] The memory stores instructions that can be executed by the processor, and the instructions are used to be executed by the processor to implement the multi-modal control and reactive voltage on-site control method of the port hybrid energy supply system as described in any of the above items.

[0163] In this embodiment, the multi-modal control and reactive voltage local control method of the port hybrid energy supply system includes the following steps: analyzing the power flow relationship of the DC microgrid under different operating conditions, and then dividing the system into working modes according to the bus voltage fluctuation range, realizing the overall coordinated control function of each unit in the system, and giving the control strategy of each unit under each operating condition in detail; using the discrete consistency algorithm in the secondary control strategy of the DC microgrid. On the basis of the primary control, two modules of voltage recovery control and current regulation control are added, and the two control objectives of DC voltage secondary recovery and load current reasonable distribution are realized at the same time. On the day-ahead time scale, according to the day-ahead forecast of photovoltaic output and load, as well as its historical statistical data, a probability model of photovoltaic output and load in each period is constructed, and the day-ahead adjustment plan of reactive output of OLTC, capacitor bank and photovoltaic converter is determined by solving an uncertain optimization model; utilizing the dynamic reactive response capability of photovoltaic inverter, adopting a real-time autonomous control strategy based on local information, and realizing the suppression of voltage over-limit and fluctuation by adjusting the reactive adjustment amount of inverter and photovoltaic active power.

Claims

1. A multi-mode control and reactive voltage local control method for a port hybrid energy supply system, characterized in that: The steps include: Step S1, analyzing the power flow relationship of the DC microgrid under different operating conditions, and dividing the system into five working modes according to the bus voltage fluctuation range; Step S2, setting two modules: voltage recovery control and current adjustment control; Step S3: Based on the day-ahead forecast of photovoltaic output and load from historical statistical data, a probability model of photovoltaic output and load in each time period is constructed to determine the day-ahead adjustment plan for the reactive output of the on-load tap changer (OLTC), capacitor bank and photovoltaic transformer; Step S4: utilizing the dynamic reactive power response capability of the photovoltaic inverter to adjust the reactive power adjustment amount of the inverter and the active power of the photovoltaic inverter to suppress voltage over-limit and fluctuation.

2. The method according to claim 1, characterized in that Step S1 is specifically as follows: Step S11, analyzing the power flow relationship of the port DC microgrid, including the power relationship under various operating conditions; Step S12: Based on the DC bus voltage control, different working modes are divided according to the power flow relationship of the port microgrid under different working conditions to achieve distributed control of energy storage, photovoltaic system and grid-connected converter, so as to achieve stable system operation.

3. The method according to claim 2, characterized in that Step S11 is specifically as follows: S111. The power balance relationship of the DC microgrid system under the grid-connected operation condition of the port DC microgrid is: P S =P PV -P bes -P load , where P S Grid exchange power, P PV is the photovoltaic power generation power, P bes is the energy storage exchange power, P load is the load power; S112, AC grid-connected converter (GCC) works in rectification and inversion mode. If the AC system and PV unit cannot meet the power requirements of the load even at full power or maximum power output, the energy storage system will provide the remaining required power and stabilize the DC bus voltage. At this time, the system enters a new stable operation state, and the power provided by the energy storage system P Bes For: P Bes =P Smax +P PV -P load , where P Smax is the maximum value of the power exchanged by the power grid; S113. If the load cannot completely consume the power generated by the distributed power source, the grid-connected converter enters the inverter working state. When the power flowing through the grid-connected converter reaches the limit value, and the sunlight is very strong or the remaining power of the energy storage system is not full, the photovoltaic and energy storage units need to jointly take over the grid-connected converter as the system's relaxation terminal to maintain the DC bus voltage. The power balance relationship at this time is: P Smax =P PV -P load -P Bes ; S114. When the port DC microgrid is in isolated operation, the energy storage system is the only balance point of the system, and the power relationship is: P Bes =P PV -P load .

4. The method according to claim 2, characterized in that The step S12 is specifically as follows: Working mode 1: U dc ≤0.95U ref The energy storage unit is responsible for maintaining the bus voltage stable and discharging to the system. The distributed power source works in MPPT mode. dc is the DC bus voltage, U ref is the rated voltage; Working mode 2: 0.95U ref ≤U dc ≤0.98U ref , when the photovoltaic power generation is in the MPPT state and is not enough to provide the power required by the load, the DC bus voltage is lower than the rated voltage, and the AC grid supplements the power shortage to the DC microgrid through the grid-connected converter at the maximum power. The energy storage unit acts as a relaxation terminal and discharges. When the SOC drops to the minimum limit, the discharge should be stopped; Working mode 3: 0.98U ref ≤U dc ≤1.02U ref At this time, the grid-connected converter and energy storage unit are used as relaxation terminals to adopt voltage droop control, the distributed power source adopts MPPT control, and the load power is provided by the distributed power source; Working mode 4: 1.02U ref ≤U dc ≤1.05U ref At this time, the DC bus voltage is higher than the rated voltage, the grid-connected converter switches to the current limiting control state, the photovoltaic unit still operates in MPPT mode, and the energy storage unit acts as a relaxation terminal to control the DC bus voltage stability; Working mode 5: U dc ≥1.05 U ref , there is too much residual power in the system, the SOC of the energy storage battery reaches the upper limit, and the photovoltaic unit adopts droop control.

5. The method according to claim 2, characterized in that In step S12, distributed control is performed on the energy storage, photovoltaic system and grid-connected converter, and step S12 is specifically as follows: Step S121: Set the droop control to be used in grid-connected and islanded operation to maintain the DC bus voltage stable. When the AC main grid fails, the microgrid needs to operate in an islanded operation for a short time. At this time, the energy storage unit will serve as a balancing node to maintain the system power balance and the stability of the DC bus voltage. Step S122, photovoltaic power generation unit control strategy, the output voltage of the photovoltaic power generation system is lower than the DC bus voltage, and the photovoltaic interface converter is connected to the DC microgrid via a BOOST type DC / DC converter; Step S123, the AC grid-connected converter is a voltage-type converter, the AC grid-connected converter is regulated by a PI regulator, the AC grid-connected converter is used to provide a DC bus voltage and maintain a power balance of the DC grid, the AC grid-connected converter adopts direct current control, the outer loop is a voltage loop, which is used to control the stability of the DC bus voltage; the inner loop is used to control the current.

6. The method according to claim 1, characterized in that The step S2 is specifically as follows: Step S21: adding distributed collaborative secondary control including voltage recovery control and current correction control to the DC microgrid control system based on discrete consistency theory; Step S22: adding a consistency gain function to reduce noise.

7. The method according to claim 6, characterized in that Step S21 is specifically as follows: Step S211, the distributed collaborative secondary control is a double closed-loop control, the double closed-loop control is a current inner loop control and a voltage outer loop return voltage droop control, the local controllers of each microgrid establish a communication connection, each controller only collects its own voltage and output current information, and only exchanges information with the controllers of adjacent microgrids; Step S212, obtaining the average voltage and average current value of each node in the entire system through consistency iteration of the information of the dual closed-loop control; Step S213, the current regulator and voltage regulator subtract the average current value and voltage value output by the consistency algorithm from the actual current value and voltage value of each node, and then amplify them through a PI controller to obtain virtual current and voltage increments to compensate for the bus voltage drop of the entire system.

8. The method according to claim 6, characterized in that The consistency gain function is: Where j is the state variable input interfered by random communication noise, and the random communication noise N j and J i It can be represented by a Gaussian white noise, which is a uniformly bounded independent random variable with a mean of 0. ij is the gain coefficient, x i (k), x j (k), v i (k) are the noises in the sending, transmission and receiving stages at time k, respectively, and x i (k+1) is the noise in the sending stage at time k+1, the consistency gain function a(k), and a(k) takes the value of ±1, ±2 or ±3.

9. The method according to claim 1, characterized in that Step S3 is specifically as follows: Step S31, reactive power optimization takes the grid voltage over-limit risk, network loss and voltage deviation as the minimum objective function, min F = λ1f1 + λ2f2 + λ3f3, where f1 is the cumulative membership function of all node voltage over-limit risks at each moment, f2 is the membership function of the expected network loss index, and f3 is the membership function of the voltage deviation index; Among them, S i,t is the voltage over-limit indicator value, V min 、V max and V B They are the maximum allowable voltage, the minimum allowable voltage and the reference voltage respectively; f i,t (x) is the voltage probability distribution of the node at time, t and T are time, i and N are nodes, x and dx are voltage and voltage differential variables; Among them, N b is the number of branches; P i,t and Q i,t are the active power and reactive power of branch i at time t respectively; R i and U i,t are the resistance of branch i and the voltage at the first end at time t respectively; Among them, V i N is the rated voltage of node i; V i,t is the voltage of node i at the moment; Fuzzy set theory is used to describe the optimization result of the objective function using the membership function F, whose value range is between 0 and 1. The smaller F is, the closer it is to the optimal value, indicating that the optimization result is more ideal. The weights λ1, λ2 and λ3 are determined using the hierarchical analysis method. Step S32, constructing constraint conditions, including: power balance constraint, energy storage processing constraint, transformer constraint, photovoltaic output constraint and photovoltaic inverter reactive output constraint; Step S33, read multiple scenario sets, each scenario set includes load and new energy output, the scenario set can be a microgrid, combine the objective function F and the constraint conditions, and output the day-ahead regulation plan of the on-load tap-changing transformer tap, capacitor bank and photovoltaic transformer reactive output; Step S4 is specifically as follows: Step S41, adjusting the active power output increment of the photovoltaic inverter according to the photovoltaic active power output prediction error; Step S42, superimposing the active power output increment of the photovoltaic inverter with the day-ahead planned value of the period to obtain the reactive power output of the inverter; Step S43: When it is detected that the voltage at the photovoltaic grid point exceeds the limit, Calculate the inverter reactive power adjustment required for PV inverter voltage regulation in, is the voltage reactive sensitivity of node i in period t, V i,τ is the grid connection point voltage value, is the voltage limit of node i.

10. A multi-mode control and reactive voltage local control method control system for a port hybrid energy supply system, characterized in that: include: at least one processor; as well as, a memory communicatively connected to at least one of the processors; wherein, The memory stores instructions executable by the processor, and the instructions are used by the processor to execute the multi-modal control and reactive voltage on-site control method of the port hybrid energy supply system as described in any one of claims 1 to 9.

Images