A distributed real-time scheduling method for isolated island microgrids adapting to power demand changes

Abstract

The present invention discloses a distributed real-time scheduling method for an isolated microgrid that adapts to changes in power demand. The method comprises: the upper-level scheduling unit of each distributed power source calculates the local output power increment in real time, and calculates the real-time optimization gradient value, then obtains the optimization gradient value of the adjacent upper-level scheduling unit through distributed communication, and then uses a distributed real-time scheduling algorithm to sequentially calculate the target optimization amount, the constraint optimization amount, and the dual variable. Finally, the obtained constraint optimization amount is output as a power reference value to the underlying control system, and the control steps are executed cyclically. The present invention can timely adjust the scheduling results according to the real-time power demand in the microgrid, track the dynamic optimal result in real time, and provide the real-time optimal output power reference value for the underlying control system of each distributed power source, thereby reducing the overall energy consumption of the microgrid and the communication frequency between the distributed power sources.

Description

Technical Field

[0001] The present invention relates to the field of microgrid distributed scheduling, and is a distributed real-time scheduling method for an island microgrid that adapts to changes in power demand. Background Art

[0002] Improving living standards and social productivity requires a secure and stable power supply. However, traditional fossil fuel power generation faces increasing shortages and environmental pollution. To address the current energy crisis and environmental protection challenges, developing and utilizing renewable energy is a key solution. However, the direct integration of large-scale distributed renewable energy into the grid poses many new challenges to grid stability. Therefore, the concept of microgrids has been proposed to achieve efficient utilization of renewable energy and provide users with a stable and reliable power supply. Typically, a microgrid consists of renewable energy generation units, dispatchable distributed power sources, energy storage devices, loads, and control devices. Depending on whether a microgrid is connected to the main grid, its operating modes can be categorized as grid-connected or islanded. In grid-connected mode, a microgrid can exchange power with the main grid to maintain internal power supply and demand balance. In contrast, microgrids operating in islanded mode require appropriate control strategies to maintain power supply and demand balance and reduce non-renewable energy consumption.

[0003] Island microgrids typically employ a hierarchical control strategy to achieve efficient and reliable operation. This strategy includes primary, secondary, and tertiary control. Primary and secondary control focus on power quality within the microgrid, while tertiary control is the scheduling layer, aiming to reduce energy consumption during microgrid operation. Due to the uncontrollable nature of renewable energy generation and load demand within a microgrid, power demand is random. Therefore, the output power of dispatchable distributed generation (DGs) is used to balance this random power demand. While primary and secondary control can achieve power supply and demand balance, the tertiary control strategy determines how to properly distribute output power among numerous DGs. Different output power ratios correspond to different energy consumption levels. Currently, microgrid scheduling methods (three-level control) are mainly divided into two types: centralized scheduling and distributed scheduling. Centralized scheduling requires a central controller to coordinate power distribution among DGs and relies heavily on communication. In contrast, distributed scheduling only requires communication between adjacent DGs to optimize energy consumption, making it more reliable and easier to implement. However, existing distributed scheduling algorithms rely on power requirements, executing multiple distributed communications and iterating the algorithm to achieve the optimal scheduling result. When power requirements change, distributed communications and iterative algorithms must be re-executed to obtain a new optimal scheduling result. This makes it impossible to track the dynamic optimal scheduling result in a timely manner, resulting in increased energy consumption and reduced distributed communication efficiency. Summary of the Invention

[0004] In response to the shortcomings of the above-mentioned existing technologies, the present invention proposes a distributed real-time scheduling method for isolated microgrids that adapts to changes in power demand, in order to timely adjust the scheduling results according to the real-time power increment of each distributed power source, and track the dynamic optimal scheduling results in real time, thereby reducing the overall energy consumption of the microgrid and improving the efficiency of distributed communication.

[0005] In order to achieve the above-mentioned object, the present invention adopts the following technical solutions:

[0006] The present invention provides a distributed real-time scheduling method for an island microgrid that adapts to power demand changes. The method is characterized in that it is applied to a scenario consisting of a plurality of distributed power sources and their corresponding underlying control systems and an upper-layer scheduling unit, and includes the following steps:

[0007] Step 1: define the current time as k and initialize k=0;

[0008] The i-th underlying control system measures the output power P of the i-th distributed generation at the current time k i k And send it to the i-th upper scheduling unit;

[0009] Define the target optimization amount of the i-th upper scheduling unit at the current k-th moment as And initialize

[0010] Define the constraint optimization quantity of the i-th upper scheduling unit at the current k-th moment as And initialize

[0011] Define the dual variable of the i-th upper scheduling unit at the current k-th moment as And initialize

[0012] Step 2: The i-th underlying control system measures the output power P of the i-th distributed power source at time k+1. i k+1 And send it to the i-th upper scheduling unit;

[0013] Step 3: The i-th upper scheduling unit calculates the output power increment at time k+1 using formula (1):

[0014]

[0015] Step 4: The i-th upper scheduling unit uses formula (2) to calculate the optimized gradient value F at time k+1 i k+1 ;

[0016]

[0017] In formula (2), f i (x) is the energy consumption function of the i-th distributed generation when the output power is x, f′ i (x) is f i The derivative of (x), ρ is the penalty factor;

[0018] Step 5: The i-th upper scheduling unit takes the optimized gradient value at time k+1 Send to all adjacent upper scheduling units, and receive the optimized gradient values ​​sent by all adjacent upper scheduling units at time k+1;

[0019] Step 6: The i-th upper scheduling unit uses formula (3) to calculate the target optimization amount at time k+1

[0020]

[0021] In formula (3), W ij is the weight coefficient between the i-th upper scheduling unit and its adjacent j-th upper scheduling unit, and W ij =W ji <0, W ji is the weight coefficient between the jth upper scheduling unit and its adjacent i-th upper scheduling unit; N i is the set of upper-layer scheduling units adjacent to the i-th upper-layer scheduling unit; is the optimized gradient value sent by the adjacent j-th upper scheduling unit at time k+1 and received by the i-th upper scheduling unit;

[0022] Step 7: The i-th upper scheduling unit uses formula (4) to calculate the constraint optimization amount at time k+1

[0023]

[0024] In formula (4), is the upper limit of the output power of the i-th distributed generation, x i is the lower limit of the output power of the i-th distributed generation;

[0025] Step 8: The i-th upper scheduling unit uses formula (5) to calculate the dual variable at time k+1

[0026]

[0027] Step 9: The i-th upper scheduling unit constrains the optimization amount The power reference value at time k+1 is output to the i-th underlying control system to control the output power of the i-th distributed generation at time k+1;

[0028] Step 10, assign k+1 to k, and return to step 2 to execute sequentially.

[0029] The electronic device of the present invention includes a memory and a processor, and is characterized in that the memory is used to store a program that supports the processor to execute the distributed real-time scheduling method of the isolated island microgrid, and the processor is configured to execute the program stored in the memory.

[0030] The present invention provides a computer-readable storage medium, wherein a computer program is stored on the computer-readable storage medium, and the computer program executes the steps of the isolated island microgrid distributed real-time scheduling method when the computer program is executed by a processor.

[0031] Compared with the prior art, the present invention has the following beneficial effects:

[0032] 1. The real-time scheduling method for the isolated island microgrid proposed in the present invention is completely distributed. By combining the alternating direction multiplier method with the distributed resource scheduling algorithm and utilizing the designed real-time power demand adaptation mechanism, it can ensure that the scheduling results are adjusted in a timely manner according to changes in power demand, so that the scheduling results closely follow the dynamic optimal results, thereby reducing the overall energy consumption of distributed power sources in the microgrid.

[0033] 2. The real-time power demand adaptation mechanism designed in the present invention uses the power increment of each distributed power source to adjust the scheduling result, so that the results of the real-time scheduling algorithm are inherited. When the power demand changes slightly, the scheduling algorithm can converge to the optimal result more quickly, so the number of communications required between adjacent distributed power sources is less, thereby improving the distributed communication efficiency in the microgrid. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 is a flow chart of the scheduling method of the present invention;

[0035] Figure 2 This is the structure diagram of the island microgrid system. DETAILED DESCRIPTION

[0036] In this embodiment, Figure 2As shown in the figure, there are several dispatchable distributed power sources in the microgrid, and each distributed power source has a corresponding underlying control system and upper-level dispatching unit. Due to the undispatching nature of renewable energy generation and load, the power supply and demand balance can only be achieved by relying on dispatchable distributed power sources. In addition, the upper-level dispatching unit needs an appropriate dispatching method to adjust the output power of the distributed power source according to the real-time power demand, in order to minimize the overall energy consumption. Therefore, microgrid scheduling can be modeled as an optimization problem as shown in formula (1);

[0037]

[0038] In formula (1), n ​​is the total number of distributed generation in the microgrid, f i (x i ) is the energy consumption function of the i-th distributed generation, x i is the output power of the i-th distributed power source, is the upper limit of the output power of the i-th distributed generation, x i is the lower bound of the output power of the ith distributed generation, and e is the real-time power demand in the microgrid;

[0039] The optimization problem shown in formula (1) is equivalent to the optimization problem shown in formula (2);

[0040]

[0041] In formula (2), is the constraint function of the i-th upper scheduling unit, z i is the constraint variable of the ith upper scheduling unit, and defines Z = [z1 … z n ] T is the constraint vector of each upper-level dispatching unit in the microgrid, X=[x1 … x n ] T is the output power vector of each distributed power source in the microgrid, is the total energy consumption function of each distributed power source in the microgrid, is the total constraint function of each upper-level dispatching unit in the microgrid, I n =[1 … 1] T is an n-dimensional column vector with all elements set to 1, then the optimization problem shown in formula (2) is equivalent to the optimization problem shown in formula (3);

[0042]

[0043] For the optimization problem shown in formula (3), the alternating direction multiplier method shown in formula (4), formula (5) and formula (6) can be used to iteratively solve it;

[0044]

[0045]

[0046] U k+1 =U k +X k+1 -Z k+1 (6)

[0047] In formula (4), formula (5) and formula (6), ρ is the penalty factor, U = [u1 … u n ] T is the dual vector of each upper dispatch unit in the microgrid, u i is the dual variable of the i-th upper scheduling unit, k refers to the number of iterations, and ||·||2 is the 2-norm; however, the iterative algorithms shown in Equations (4), (5), and (6) are centralized algorithms and need to rely on a centralized controller, so Equations (4), (5), and (6) need to be designed as distributed algorithms; then, each distributed power source can obtain the optimal scheduling result according to the distributed iterative algorithm;

[0048] Specifically, if Figure 1 As shown, a distributed real-time scheduling method for an island microgrid that adapts to changes in power demand includes the following steps:

[0049] Step 1: define the current time as k and initialize k=0;

[0050] The i-th underlying control system measures the output power P of the i-th distributed generation at the current time k i k And send it to the i-th upper scheduling unit;

[0051] Define the target optimization amount of the i-th upper scheduling unit at the current k-th moment as And initialize

[0052] Define the constraint optimization quantity of the i-th upper scheduling unit at the current k-th moment as And initialize

[0053] Define the dual variable of the i-th upper scheduling unit at the current k-th moment as And initialize

[0054] Using the distributed resource scheduling algorithm, Equation (4) can be improved into a distributed iterative algorithm as shown in Equation (5);

[0055]

[0056] In formula (7), W ij is the weight coefficient between the i-th upper scheduling unit and its adjacent j-th upper scheduling unit; N i is the set of upper-level scheduling units adjacent to the i-th upper-level scheduling unit; in order to adapt to the change of real-time power demand e in the microgrid, a real-time power demand adaptation mechanism as shown in steps 2 to 6 is designed;

[0057] Step 2: The i-th underlying control system measures the output power P of the i-th distributed power source at time k+1. i k+1 And send it to the i-th upper scheduling unit;

[0058] Step 3: The i-th upper scheduling unit calculates the output power increment at time k+1 using formula (8):

[0059]

[0060] Step 4: The i-th upper scheduling unit uses formula (9) to calculate the optimized gradient value F at time k+1 i k+1 ;

[0061]

[0062] In formula (9), f i (x) is the energy consumption function of the i-th distributed generation when the output power is x, f′ i (x) is f i The derivative of (x), ρ is the penalty factor;

[0063] Step 5, such as Figure 2 As shown in the figure, in the distributed scheduling method, no centralized controller is required, and only the communication between adjacent upper-level scheduling units is relied upon; the i-th upper-level scheduling unit takes the optimized gradient value F at time k+1 as i k+1 Send to all adjacent upper scheduling units, and receive the optimized gradient values ​​sent by all adjacent upper scheduling units at time k+1;

[0064] Step 6: At this time, Equation (7) is improved into a distributed iterative algorithm as shown in Equation (10), which can ensure that the iterative result can adapt to the changes in the real-time power demand e and converge to the dynamic optimal result; the i-th upper scheduling unit uses Equation (10) to calculate the target optimization quantity at time k+1

[0065]

[0066] In formula (10), Wij is the weight coefficient between the i-th upper scheduling unit and its adjacent j-th upper scheduling unit, and W ij =W ji <0, W ji is the weight coefficient between the jth upper scheduling unit and its adjacent i-th upper scheduling unit; N i is the set of upper-layer scheduling units adjacent to the i-th upper-layer scheduling unit; is the optimized gradient value sent by the adjacent j-th upper scheduling unit at time k+1 and received by the i-th upper scheduling unit;

[0067] Step 7: Implement Equation (5) in a distributed manner to obtain the distributed iterative algorithm shown in Equation (11); the i-th upper scheduling unit uses Equation (11) to calculate the constraint optimization quantity at time k+1

[0068]

[0069] In formula (11), is the upper limit of the output power of the ith distributed power source, x i is the lower limit of the output power of the i-th distributed generation;

[0070] Step 8: Implement Equation (6) in a distributed manner to obtain the distributed iterative algorithm shown in Equation (12); the i-th upper scheduling unit uses Equation (12) to calculate the dual variable at time k+1

[0071]

[0072] Step 9: The i-th upper scheduling unit constrains the optimization amount The power reference value at time k+1 is output to the i-th underlying control system to control the output power of the i-th distributed generation at time k+1;

[0073] Step 10, such as Figure 1 As shown, assign k+1 to k and return to step 2 to execute sequentially.

[0074] In this embodiment, an electronic device includes a memory and a processor, wherein the memory is used to store a program that supports the processor to execute the above method, and the processor is configured to execute the program stored in the memory.

[0075] In this embodiment, a computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above method are executed.

[0076] In summary, this method can adjust the scheduling results in time according to the real-time power demand in the microgrid, track the dynamic optimal results in real time, and provide real-time optimal output power reference values ​​for the underlying control systems of each distributed power source, thereby reducing the overall energy consumption of the microgrid. In addition, the real-time scheduling algorithm adjusts the scheduling results according to the output power increment, so that the optimization results are inherited and can effectively reduce the communication frequency between distributed power sources.

Claims

1. A distributed real-time scheduling method for an island microgrid that adapts to power demand changes, characterized in that: It is applied to a scenario consisting of several distributed power sources and their corresponding underlying control systems and upper-level dispatching units, and includes the following steps: Step 1: define the current time as k and initialize k=0; The i-th underlying control system measures the output power P of the i-th distributed generation at the current time k i k And send it to the i-th upper scheduling unit; Define the target optimization amount of the i-th upper scheduling unit at the current k-th moment as And initialize Define the constraint optimization quantity of the i-th upper scheduling unit at the current k-th moment as And initialize Define the dual variable of the i-th upper scheduling unit at the current k-th moment as And initialize Step 2: The i-th underlying control system measures the output power P of the i-th distributed power source at time k+1. i k+1 And send it to the i-th upper scheduling unit; Step 3: The i-th upper scheduling unit calculates the output power increment at time k+1 using formula (1): Step 4: The i-th upper scheduling unit uses formula (2) to calculate the optimized gradient value F at time k+1 i k+1 ; In formula (2), f i (x) is the energy consumption function of the i-th distributed generation when the output power is x, f i ′(x) is f i The derivative of (x), ρ is the penalty factor; Step 5: The i-th upper scheduling unit takes the optimized gradient value F at time k+1 i k+1 Send to all adjacent upper scheduling units, and receive the optimized gradient values ​​sent by all adjacent upper scheduling units at time k+1; Step 6: The i-th upper scheduling unit uses formula (3) to calculate the target optimization amount at time k+1 In formula (3), W ij is the weight coefficient between the i-th upper scheduling unit and its adjacent j-th upper scheduling unit, and W ij =W ji <0, W ji is the weight coefficient between the jth upper scheduling unit and its adjacent i-th upper scheduling unit; N i is the set of upper-layer scheduling units adjacent to the i-th upper-layer scheduling unit; is the optimized gradient value sent by the adjacent j-th upper scheduling unit at time k+1 and received by the i-th upper scheduling unit; Step 7: The i-th upper scheduling unit uses formula (4) to calculate the constraint optimization amount at time k+1 In formula (4), is the upper limit of the output power of the ith distributed power source, x i is the lower limit of the output power of the i-th distributed generation; Step 8: The i-th upper scheduling unit uses formula (5) to calculate the dual variable at time k+1 Step 9: The i-th upper scheduling unit constrains the optimization amount The power reference value at time k+1 is output to the i-th underlying control system to control the output power of the i-th distributed generation at time k+1; Step 10, assign k+1 to k, and return to step 2 to execute sequentially.

2. An electronic device comprising a memory and a processor, characterized in that: The memory is used to store a program that supports the processor to execute the distributed real-time scheduling method for the isolated island microgrid according to claim 1, and the processor is configured to execute the program stored in the memory.

3. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the isolated island microgrid distributed real-time scheduling method according to claim 1 are executed.

Images