A method and system for dispatching loads in a park grid-connected operation

Abstract

The present invention relates to a load dispatching method and system for grid-connected operation in a park, belonging to the technical field of electric power systems, and comprising: establishing a load dispatching model, including a first dispatching target and pre-set restriction conditions; based on the electricity price information of industrial loads and domestic loads in peak and valley time periods and the disconnection tolerance conditions, generating a second dispatching target to update the objective function, and configuring corresponding time compensation values ​​for the predicted power of industrial loads and the predicted power of domestic loads, respectively, and solving the global optimal solution of the load dispatching model by using a particle swarm algorithm and a firefly algorithm as well as a set population exchange strategy. Beneficial effects: Taking full account of the power consumption characteristics of industrial loads and domestic loads in the park, by introducing time compensation values, the power consumption guarantee and cost optimization effect under the condition of multi-energy power supply are guaranteed, and a power dispatching scheme with more practical application significance is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of power systems, and in particular to a method and system for load dispatching for grid-connected operation of a park. Background Art

[0002] With the development of social economy and the progress of science and technology, the increasing exhaustion of traditional energy, the pressure of coal-based energy structure and the deteriorating ecological environment have attracted more and more attention. In order to cope with this serious problem, people have increased the research and development of renewable energy. Photovoltaic microgrids based on solar energy have gradually become the research direction of renewable energy. How to coordinate the dispatch of photovoltaic generators, energy storage systems and loads to reduce the cost of grid-connected operation is of great research significance.

[0003] The existing coordinated optimization dispatching scheme mainly focuses on how to arrange the output of each unit under grid-connected operation, optimally distribute the load and coordinate the purchase and sale of electricity by the grid, but does not take into account the different types of loads in the grid-connected park and the change rules and error tolerance of different types of loads, which leads to differences in the optimization scheme. Therefore, it is necessary to study the dispatching method for different characteristics of loads. Summary of the invention

[0004] In order to solve the above technical problems, the present invention provides a method and system for load scheduling of a park grid-connected operation.

[0005] The technical problem solved by the present invention can be achieved by adopting the following technical solutions:

[0006] A method for dispatching loads in a park grid-connected operation, comprising:

[0007] Obtaining industrial load forecast power, domestic load forecast power and photovoltaic forecast power corresponding to a scheduling time period;

[0008] Establishing a load dispatching model in a grid-connected operation state, wherein the load dispatching model includes a first dispatching target and a preset constraint condition;

[0009] Based on the electricity price information of industrial load and domestic load in the park during the peak and valley time periods and the power disconnection tolerance conditions, a second scheduling target is generated to update the objective function of the load scheduling model, and corresponding time compensation values ​​are configured for the industrial load forecast power and the domestic load forecast power, respectively, and the particle swarm algorithm and the firefly algorithm and the set population exchange strategy are used to solve the global optimal solution of the load scheduling model;

[0010] The photovoltaic power generation group, energy storage system, power consumption system and the exchange energy between the external power grid in the park are configured according to the global optimal solution to realize the load scheduling of the grid-connected operation system.

[0011] Preferably, the first scheduling target is:

[0012]

[0013] Among them, the objective function is to minimize the power consumption cost C of the power grid under the grid-connected operation state; Q(t) is the power consumption cost at time t; P n (t) is the power supply power of the power grid at time t.

[0014] Preferably, the restriction conditions include power balance restriction, load balance restriction, grid power restriction, photovoltaic power generation power restriction, energy storage power supply power restriction, ramp rate restriction, and conversion restriction;

[0015] The power balance limit is:

[0016]

[0017] Among them, P n (t) is the power supply of the power grid at time t; P p.i (t) is the power supply of the ith photovoltaic power generation group at time t; Np is the number of photovoltaic power generation groups; P c.j (t) is the power supply power of the jth energy storage system at time t; Nc is the number of energy storage systems; P load (t) is the total load of the power system at time t;

[0018] The load balancing limitations are:

[0019] P load (t) = P load工 (t)+P load居 (t);

[0020] Among them, P load工 (t) is the predicted power of the industrial load at time t; P load居 (t) is the predicted power of the life load at time t; P load (t) is the total load of the power system at time t;

[0021] The grid power limit:

[0022]

[0023] Among them, P n (t) is the power supply of the power grid at time t; Minimum power constraints for supplying power to the grid; Maximum power constraints for supplying power to the grid;

[0024] The photovoltaic power generation power limit is:

[0025]

[0026] Among them, P p.i (t) is the power generation of the ith photovoltaic power generation group at time t; is the minimum power constraint of the i-th photovoltaic power generation group; is the maximum power constraint of the i-th photovoltaic power generation group;

[0027] The energy storage power supply power limit is:

[0028] P c.j (t) = V j (t)-V j (t-1)

[0029]

[0030] Among them, P c.j (t) is the power supply power of the jth energy storage system at time t; V j (t) is the energy storage capacity of the j-th energy storage system at time t; V j (t-1) is the energy storage capacity of the j-th energy storage system at time t-1; is the maximum capacity of the j-th energy storage system;

[0031] The ramp rate limits are:

[0032] Down n ≤P n (t)-P n (t-1)≤Up n

[0033] Down p.i ≤P p.i (t)-P p.i (t-1)≤Up p.i

[0034] Down c.j ≤P c.j (t)-P c.j (t-1)≤Up c.j ;

[0035] Among them, Down n Up is the limit of the downward climbing rate of the power grid; n P is the upward climbing rate limit of the power grid; n (t) is the power supply of the power grid at time t; P n (t-1) is the power supply of the power grid at time t-1; Down p.iis the downward climbing rate limit of the ith photovoltaic power generation group; Up p.i is the upward climbing rate limit of the ith photovoltaic power generation group; P p.i (t), P p.i (t-1) is the power generation of the ith photovoltaic power generation group at time t and time t-1; Down c.j Up is the downward climbing rate limit of the j-th energy storage system; c.j is the upward climbing rate limit of the j-th energy storage system; P c.j (t), P c.j (t-1) is the power supply power of the j-th energy storage system at time t and time t-1.

[0036] Conversion restrictions:

[0037] P p.i (t) = η p.i E p.i (t)

[0038] P c.j (t) = η c.j U c.j (t);

[0039] Among them, E p.i (t) is the predicted photovoltaic power of the i-th photovoltaic power generation group at time t; U c.j (t) is the theoretical output power of the j-th energy storage system at time t; η p.i , η c.j is the conversion rate parameter.

[0040] Preferably, the second scheduling target includes:

[0041]

[0042] Among them, the objective function is to minimize the power consumption cost F of the power grid under the grid-connected operation state; Q 工 (t) is the industrial electricity cost at time t; P load工 (t) is the predicted power of the industrial load at time t; Q 居 (t) is the electricity cost of life at time t; P load居 (t) is the predicted power of the living load at time t.

[0043] Preferably, corresponding time compensation values ​​are configured for the industrial load forecast power and the life load forecast power respectively:

[0044] P load (t) = βP load工 (f)+γP load居 (t);

[0045] Among them, P load工 (t) is the predicted power of the industrial load at time t; β is the time compensation value corresponding to the industrial load; P load居 (t) is the predicted power of the life load at time t; γ is the time compensation value corresponding to the life load; P load (t) is the total load of the power system at time t.

[0046] Preferably, the method of solving the global optimal solution of the load scheduling model by using the particle swarm algorithm and the firefly algorithm and a set population exchange strategy includes:

[0047] Using a particle swarm algorithm, the number, position and speed of the initialized particles are determined according to the constraint conditions, and the first fitness value of each particle is calculated according to the scheduling target, and the position is updated;

[0048] Using the firefly algorithm, the initialization maximum attraction, attraction factor, step size, and initial position are determined according to the constraints, and the number of individuals, population number, and dimension that match the number of initialized particles of the particle swarm algorithm are determined according to the constraints, and the second fitness value of each individual is calculated according to the scheduling target, and the position is updated;

[0049] When the evolutionary generations of the particle swarm algorithm and the firefly algorithm reach a preset generation threshold, the first fitness value and the second fitness value are sorted respectively, and exchanged according to a set population exchange strategy; otherwise, the iteration continues;

[0050] Recording the global optimal values ​​of the particle swarm algorithm and the firefly algorithm;

[0051] Determine whether the loop stop condition is reached:

[0052] If it is reached, the iteration process is exited and the global optimal value is output; otherwise, the iteration process continues;

[0053] Preferably, the set population exchange strategy includes:

[0054] Step A1, determine whether it is the first exchange:

[0055] If yes, proceed to step A2;

[0056] If not, proceed to step A3;

[0057] Step A2, at the first preset number of evolutionary generations, the results of the first N individuals, the middle N individuals and the last N individuals of the fitness value are respectively exchanged, where N is a positive integer;

[0058] Step A3, at the second preset number of evolutionary generations, respectively take the results of the first M individuals and the last M individuals of the fitness value and exchange them, where M is a positive integer.

[0059] Preferably, in step A2, the value of N increases with the increase of the current evolution generation.

[0060] Preferably, in step A2, the value of N is determined using the following formula:

[0061]

[0062] Among them, z is the current evolutionary generation; S represents the population size; k represents the control coefficient; z represents the current evolutionary generation; Zmax represents the maximum evolutionary generation.

[0063] The present invention also provides a park grid-connected operation load dispatching system, which is applied to the park grid-connected operation load dispatching method as described above, comprising:

[0064] An acquisition unit, used for acquiring industrial load forecast power, domestic load forecast power and photovoltaic forecast power corresponding to a scheduling time period by using a preset forecast model;

[0065] A model building unit, used to build a load dispatching model in a grid-connected operation state, wherein the load dispatching model includes a dispatching target and pre-set restriction conditions, wherein the dispatching target is to minimize the grid electricity cost in the grid-connected operation state, and the restriction conditions include power balance restriction, load balance restriction, grid power restriction, photovoltaic power generation power restriction, energy storage power supply power restriction, ramp rate restriction, and conversion restriction;

[0066] An optimization unit is used to generate a second scheduling target to update the objective function of the load scheduling model based on the electricity price information of the industrial load and the domestic load in the park during the peak and valley time periods and the power disconnection tolerance conditions, and configure corresponding time compensation values ​​for the industrial load forecast power and the domestic load forecast power respectively, and use the particle swarm algorithm and the firefly algorithm and the set population exchange strategy to solve the global optimal solution of the load scheduling model;

[0067] The dispatching unit is used to reasonably configure the exchange energy between the photovoltaic power generation group, energy storage system, power consumption system and external power grid in the park according to the global optimal solution, so as to realize the load dispatching of the grid-connected operation system.

[0068] The advantages or beneficial effects of the technical solution of the present invention are:

[0069] The present invention aims at the electricity consumption characteristics of industrial loads and domestic loads in the park, fully considers the electricity price information of loads with different characteristics during peak and valley time periods and their power outage tolerance conditions, and introduces a time compensation value to ensure power consumption and cost optimization under multi-energy power supply conditions, thereby realizing an electric energy scheduling solution with more practical application significance. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] Figure 1 A flow chart of a load dispatching method for grid-connected operation of a park in a preferred embodiment of the present invention;

[0071] Figure 2 This is a flow chart of solving the load scheduling model using the particle swarm algorithm and the firefly algorithm in a preferred embodiment of the present invention;

[0072] Figure 3 A schematic diagram of a flow chart of a population exchange strategy set in a preferred embodiment of the present invention;

[0073] Figure 4 This is a schematic diagram of the overall system of a grid-connected park in a preferred embodiment of the present invention;

[0074] Figure 5 This is a structural block diagram of the load dispatching system for grid-connected operation in the Shen Park in a preferred embodiment of the present invention. DETAILED DESCRIPTION

[0075] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0076] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments may be combined with each other.

[0077] The present invention will be further described below with reference to the accompanying drawings and specific embodiments, but they are not intended to limit the present invention.

[0078] The method and system for load scheduling of a park grid-connected operation in the embodiment of the present invention are applicable to a grid-connected operation park. Figure 4 As shown, the grid-connected operation system in the park includes a photovoltaic power generation system 1, an energy storage station 2, a power consumption system 3 and an externally connected power grid 4, wherein the photovoltaic power generation system 1 has at least two photovoltaic power generation groups, the energy storage station 2 has at least two energy storage systems, and the power consumption system 3 is divided into industrial electricity and domestic electricity.

[0079] The photovoltaic power generation system 1 is equipped with a photovoltaic data acquisition unit for collecting photovoltaic data and storing it in a database for the neural network model algorithm to predict photovoltaic power. The power consumption system 3 is equipped with an industrial load acquisition unit and a life load acquisition unit. The industrial load acquisition unit is used to collect industrial load data in the park and store it in a database for the neural network model algorithm to predict industrial load power; the life load acquisition unit is used to collect life load data in the park and store it in a database for the neural network model algorithm to predict life load power.

[0080] In a preferred embodiment of the present invention, based on the above problems existing in the prior art, a method for load dispatching of a park grid-connected operation is provided, which belongs to the technical field of power system. Figure 1 As shown, including:

[0081] Obtaining industrial load forecast power, domestic load forecast power and photovoltaic forecast power corresponding to a scheduling time period;

[0082] Specifically, in this embodiment, the industrial load prediction power, the life load prediction power and the photovoltaic prediction power can be predicted by a preset prediction model. For example, the preset prediction model may include an industrial load prediction LSTM neural network model, a life load prediction LSTM neural network model, and a photovoltaic prediction LSTM neural network model. The industrial load prediction power is predicted by the industrial load prediction LSTM neural network model; the life load prediction power is predicted by the life load prediction LSTM neural network model; and the photovoltaic prediction power is predicted by the photovoltaic prediction LSTM neural network model.

[0083] Preferably, the above-mentioned industrial load prediction LSTM neural network model, life load prediction LSTM neural network model and photovoltaic prediction LSTM neural network model can all adopt existing network models, which will not be repeated here.

[0084] Establishing a load dispatching model in a grid-connected operation state, the load dispatching model including a first dispatching target and a preset constraint condition;

[0085] As a preferred implementation, the first scheduling target is:

[0086]

[0087] Among them, the objective function is to minimize the power consumption cost C of the power grid under the grid-connected operation state; Q(t) is the power consumption cost at time t; P n (t) is the power supply power of the power grid at time t.

[0088] The above-mentioned first dispatching target is to ensure the minimum grid electricity cost while ensuring uninterrupted power supply for industrial and domestic use.

[0089] As a preferred implementation, the restriction conditions include power balance restriction, load balance restriction, grid power restriction, photovoltaic power generation power restriction, energy storage power supply power restriction, ramp rate restriction, and conversion restriction;

[0090] Power balance limits:

[0091]

[0092] Among them, P n (t) is the power supply of the power grid at time t; P p.i (t) is the power supply of the i-th photovoltaic power generation group at time t; Np is the number of photovoltaic power generation groups; P c.j (t) is the power supply of the jth energy storage system at time t; Nc is the number of energy storage systems; P load (t) is the total load of the power system at time t, which includes industrial load and domestic load;

[0093] Load Balancing Limitations:

[0094] P load (t) = P load工 (t)+P load居 (t);

[0095] Among them, P load工 (t) is the predicted industrial load power at time t; P load居 (t) is the predicted power of life load at time t; P load (t) is the total load of the power system at time t;

[0096] Grid power limitation:

[0097]

[0098] Among them, P n (t) is the power supply of the power grid at time t; Minimum power constraints for supplying power to the grid; Maximum power constraints for supplying power to the grid;

[0099] Photovoltaic power generation power limit:

[0100]

[0101] Among them, P p.i (t) is the power generation of the i-th photovoltaic generator set at time t; is the minimum power constraint of the i-th photovoltaic generator set; is the maximum power constraint of the i-th photovoltaic generator set;

[0102] Energy storage power supply power limit:

[0103] P c.j (t) = V j (t)-V j (t-1)

[0104]

[0105] Among them, P c.j (t) is the power supply of the jth energy storage system at time t; V j (t) is the energy storage capacity of the jth energy storage system at time t; V j (t-1) is the energy storage capacity of the jth energy storage system at time t-1; is the maximum capacity of the jth energy storage system;

[0106] Ramp rate limit:

[0107] Down n ≤P n (t)-P n (t-1)≤Up n

[0108] Down p.i ≤P p.i (t)-P p.i (t-1)≤Up p.i

[0109] Down c.j ≤P c.j (t)-P c.j (t-1)≤Up c.j ;

[0110] Among them, Down n Up is the limit of the downward climbing rate of the power grid; n P is the upward climbing rate limit of the power grid; n (t) is the power supply of the power grid at time t; P n (t-1) is the power supply of the power grid at time t-1; Down p.i is the downward climbing rate limit of the i-th photovoltaic generator set; Up p.i is the upward climbing rate limit of the i-th photovoltaic generator set; P p.i (t), P p.i (t-1) is the power generation of the ith photovoltaic generator set at time t and time t-1; Down c.j Up is the downward climbing rate limit of the j-th energy storage system; c.j is the upward climbing rate limit of the j-th energy storage system; P c.j (t), Pc.j (t-1) is the power supply power of the j-th energy storage system at time t and time t-1.

[0111] The above photovoltaic power generation group and energy storage system have certain energy conversion rate limitations and energy losses, so η is set respectively p.i , η c.j As a conversion rate parameter, the conversion limit:

[0112] P p.i (t) = η p.i E p.i (t)

[0113] P c.j (t) = η c.j U c.j (t);

[0114] Among them, E p.i (t) is the predicted photovoltaic power of the i-th photovoltaic generator set at time t; U c.j (t) is the theoretical output power of the jth energy storage system at time t; η p.i , η c.j is the conversion rate parameter.

[0115] Based on the electricity price information of industrial load and domestic load in the park during the peak and valley time periods and the power outage tolerance conditions, the second scheduling target is generated to update the objective function of the load scheduling model, and the corresponding time compensation values ​​are configured for the industrial load forecast power and the domestic load forecast power respectively. The particle swarm algorithm and the firefly algorithm as well as the set population exchange strategy are used to solve the global optimal solution of the load scheduling model.

[0116] As a preferred implementation, the second scheduling target includes:

[0117]

[0118] Among them, the objective function is to minimize the power consumption cost F of the power grid under the grid-connected operation state; Q 工 (t) is the industrial electricity cost at time t; P load工 (t) is the predicted industrial load power at time t; Q 居 (t) is the electricity cost of life at time t; P load居 (t) is the predicted power of life load at time t.

[0119] Specifically, the electricity costs of industrial electricity and domestic electricity are different, and the electricity costs of domestic electricity during peak and valley time periods are also different. The above-mentioned second scheduling target comprehensively considers the characteristics of industrial load and domestic load as well as the electricity price information during peak and valley time periods to ensure the minimum grid electricity cost while meeting the requirements of power outage tolerance of industrial load and domestic load in different time periods.

[0120] As a preferred implementation, corresponding time compensation values ​​are configured for the industrial load forecast power and the life load forecast power respectively:

[0121] P load (f) = βP load工 (t)+γP load居 (t);

[0122] Among them, P load工 (t) is the predicted power of industrial load at time t; β is the time compensation value corresponding to industrial load; P load居 (t) is the predicted power of the life load at time t; γ is the time compensation value corresponding to the life load; P load (t) is the total load of the power system at time t.

[0123] Specifically, since industrial load and life load have different change rules and error tolerances, for example, industrial load occupies a major position in the power consumption of the park, its load is large, and the demand is relatively stable. At the same time, power outages have a greater impact on industrial power consumption. Therefore, in the scheduling process, it is necessary to ensure that industrial power is almost uninterrupted. The characteristics of life load are that it is greatly affected by the work and rest of the park, weather conditions and electricity usage habits, and the periodic fluctuations are more obvious. The periodicity includes daily cycles and annual cycles. The life load is small during the working time period (for example, 8:00 to 20:00), and the load is large during the night time period (for example, 20:00 to 8:00 the next day). The load demand in winter and summer is larger than that in spring and autumn.

[0124] In view of the above characteristics, different restrictions will be imposed on the input end when establishing the load scheduling model, that is, the industrial load forecast power and the domestic load forecast power have different time compensation values. The time compensation value is variable. The specific value of the time compensation value is related to the disconnection allowance of the industrial load and the domestic load. The specific value can be set according to actual conditions.

[0125] For example, including but not limited to the following examples: During working hours (e.g., 8:00 to 20:00), P load (t) = 1.1P load工 (t)+P load居 (t), that is, β = 1.1, γ = 1;

[0126] In the night time period (e.g. 20:00 to 8:00 the next day), it can be specifically divided into spring and autumn seasons and winter and summer seasons;

[0127] In spring and autumn: P load (t) = 1P load工 (t)+1.05P load居 (t)

[0128] In winter and summer: P load (t) = 1P load工 (t)+1.1P load居 (t).

[0129] The energy exchange between the photovoltaic power generation units, energy storage systems, power consumption systems and external power grids in the park is configured according to the global optimal solution to achieve load scheduling of the grid-connected operation system.

[0130] Specifically, the embodiments of the present invention fully consider the electricity consumption characteristics of industrial loads and domestic loads, as well as the electricity price information and power outage tolerance during peak and valley periods, and perform scheduling control based on the predicted power of industrial loads, predicted power of domestic loads and predicted power of photovoltaic power to minimize the power supply cost of the park power grid and ensure stable power supply under different park time conditions.

[0131] As a preferred implementation, the particle swarm algorithm and the firefly algorithm as well as the set population exchange strategy are used to solve the global optimal solution of the load scheduling model, such as Figure 2 As shown, including:

[0132] The particle swarm algorithm is used to determine the number, position and speed of the initialized particles according to the constraints. At the same time, the first fitness value of each particle is calculated according to the scheduling target, and the position is updated;

[0133] The firefly algorithm is used to determine the initial maximum attraction, attraction factor, step size, and initial position according to the constraints. At the same time, the number of individuals, population number, and dimension that match the number of initial particles of the particle swarm algorithm are determined according to the constraints. The second fitness value of each individual is calculated according to the scheduling target, and the position is updated;

[0134] When the evolutionary generations of the particle swarm algorithm and the firefly algorithm reach the preset generation threshold, the first fitness value and the second fitness value are sorted respectively and exchanged according to the set population exchange strategy; otherwise, the iteration continues;

[0135] Record the global optimal values ​​of particle swarm algorithm and firefly algorithm;

[0136] Determine whether the loop stop condition is reached:

[0137] If it is reached, the iteration process is exited and the global optimal value is output; otherwise, the iteration process continues;

[0138] As a preferred embodiment, wherein Figure 3 As shown, the population exchange strategy set includes:

[0139] Step A1, determine whether it is the first exchange:

[0140] If yes, proceed to step A2;

[0141] If not, proceed to step A3;

[0142] Step A2, at the first preset number of evolutionary generations, the results of the first N individuals, the middle N individuals and the last N individuals of the fitness value are respectively exchanged, where N is a positive integer;

[0143] Step A3, at the second preset number of evolution generations, respectively take the results of the first M individuals and the last M individuals of the fitness value and exchange them, where M is a positive integer.

[0144] As a preferred implementation, in step A2, the value of N increases with the increase of the current evolutionary generation.

[0145] As a preferred implementation, in step A2, the value of N is determined using the following formula:

[0146]

[0147] Among them, z is the current evolutionary generation; S represents the population size; k represents the control coefficient; z represents the current evolutionary generation; Zmax represents the maximum evolutionary generation.

[0148] Furthermore, the inertia weight is updated according to the fitness value, and the updating method specifically includes:

[0149] Compare the individual's fitness value with the preset error value to determine whether the individual's fitness value is greater than the preset error value:

[0150] If the fitness value of the individual is greater than the preset error value, that is, when it is far away from the local optimal solution, the inertia weight is updated according to the first preset update principle;

[0151] If the fitness value of the individual is less than or equal to the preset error value, that is, when it is close to the local optimal solution, the inertia weight is updated according to the second preset update principle.

[0152] Specifically, the first preset update principle includes:

[0153]

[0154] Among them, α' is the reference value of α; λ' is the compensation coefficient; v kd is the velocity of the particle.

[0155] Specifically, the second preset update principle includes:

[0156]

[0157] Among them, α' is the reference value of α; λ" is the compensation coefficient; v kd is the velocity of the particle.

[0158] Specifically, in this embodiment, the individual exchange through the population exchange strategy of the above two algorithms can effectively avoid local optimality, but it also slows down the optimization speed. Therefore, the embodiment of the present invention optimizes the convergence effect of the algorithm by setting inertia weights, and its retrieval is more detailed, making the local search more accurate.

[0159] The present invention also provides a park grid-connected operation load dispatching system, which is applied to the park grid-connected operation load dispatching method as described above. Figure 5 As shown, including:

[0160] An acquisition unit 5 is used to acquire industrial load forecast power, domestic load forecast power and photovoltaic forecast power corresponding to a scheduling time period by using a preset forecast model;

[0161] The model building unit 6 is used to build a load dispatching model in the grid-connected operation state. The load dispatching model includes a dispatching target and pre-set constraints. The dispatching target is to minimize the grid electricity cost in the grid-connected operation state. The constraints include power balance constraint, load balance constraint, grid power constraint, photovoltaic power generation power constraint, energy storage power supply power constraint, ramp rate constraint, and conversion constraint.

[0162] The optimization unit 7 is used to generate a second scheduling target to update the objective function of the load scheduling model based on the electricity price information of the industrial load and the domestic load in the park during the peak and valley time periods and the disconnection tolerance conditions, and configure corresponding time compensation values ​​for the industrial load forecast power and the domestic load forecast power respectively, and use the particle swarm algorithm and the firefly algorithm and the set population exchange strategy to solve the global optimal solution of the load scheduling model;

[0163] The dispatching unit 8 is used to reasonably configure the exchange energy between the photovoltaic power generation group, energy storage system, power consumption system and the external power grid in the park according to the global optimal solution, so as to realize the load dispatching of the grid-connected operation system.

[0164] The adoption of the above technical solution has the following advantages or beneficial effects: the present invention targets the electricity consumption characteristics of industrial loads and domestic loads in the park, fully considers the electricity price information of loads with different characteristics during peak and valley time periods and their power outage tolerance conditions, and introduces a time compensation value to ensure electricity consumption and cost optimization under multi-energy power supply conditions, thereby realizing an electric energy scheduling solution with more practical application significance.

[0165] The above description is only a preferred embodiment of the present invention, and does not limit the implementation mode and protection scope of the present invention. For those skilled in the art, it should be aware that all solutions obtained by equivalent substitutions and obvious changes made using the contents of this specification and illustrations should be included in the protection scope of the present invention.

Claims

1. A method for load dispatching of a park grid-connected operation, characterized in that: include: Obtaining industrial load forecast power, domestic load forecast power and photovoltaic forecast power corresponding to a scheduling time period; Establishing a load dispatching model in a grid-connected operation state, wherein the load dispatching model includes a first dispatching target and a preset constraint condition; Based on the electricity price information of industrial load and domestic load in the park during the peak and valley time periods and the power disconnection tolerance conditions, a second scheduling target is generated to update the objective function of the load scheduling model, and corresponding time compensation values ​​are configured for the industrial load forecast power and the domestic load forecast power, respectively, and the particle swarm algorithm and the firefly algorithm and the set population exchange strategy are used to solve the global optimal solution of the load scheduling model; According to the global optimal solution, the photovoltaic power generation group, energy storage system, power consumption system and the exchange energy between the external power grid in the park are configured to realize the load scheduling of the grid-connected operation system; The second scheduling target includes: Among them, the objective function is to minimize the power consumption cost F of the power grid under the grid-connected operation state; Q 工 (t) is the industrial electricity cost at time t; P load工 (t) is the predicted power of the industrial load at time t; Q 居 (t) is the electricity cost of life at time t; P load居 (t) is the predicted power of the life load at time t; The industrial load forecast power and the life load forecast power are respectively configured with corresponding time compensation values: ； Among them, P load工 (t) is the predicted power of the industrial load at time t; β is the time compensation value corresponding to the industrial load; P load居 (t) is the predicted power of the life load at time t; γ is the time compensation value corresponding to the life load; P load (t) is the total load of the power system at time t; The method of using the particle swarm algorithm and the firefly algorithm as well as the set population exchange strategy to solve the global optimal solution of the load scheduling model includes: Using a particle swarm algorithm, the number, position and speed of the initialized particles are determined according to the constraint conditions, and the first fitness value of each particle is calculated according to the scheduling target, and the position is updated; Using the firefly algorithm, the initialization maximum attraction, attraction factor, step size, and initial position are determined according to the constraints, and the number of individuals, population number, and dimension that match the number of initialized particles of the particle swarm algorithm are determined according to the constraints, and the second fitness value of each individual is calculated according to the scheduling target, and the position is updated; When the evolutionary generations of the particle swarm algorithm and the firefly algorithm reach a preset generation threshold, the first fitness value and the second fitness value are sorted respectively, and exchanged according to a set population exchange strategy; otherwise, the iteration continues; Recording the global optimal values ​​of the particle swarm algorithm and the firefly algorithm; Determine whether the loop stop condition is reached: If it is reached, the iteration process is exited and the global optimal value is output; otherwise, the iteration process continues.

2. The method for load dispatching of park grid-connected operation according to claim 1, characterized in that: The first scheduling target is: ； Among them, the objective function is to minimize the power consumption cost C of the power grid under the grid-connected operation state; Q(t) is the power consumption cost at time t; P n (t) is the power supply power of the power grid at time t.

3. The method for load dispatching of a park grid-connected operation according to claim 1, characterized in that: The restriction conditions include power balance restriction, load balance restriction, grid power restriction, photovoltaic power generation power restriction, energy storage power supply power restriction, ramp rate restriction, and conversion restriction; The power balance limit is: Among them, P n (t) is the power supply of the power grid at time t; P p.i (t) is the power supply of the ith photovoltaic power generation group at time t; Np is the number of photovoltaic power generation groups; P c.j (t) is the power supply of the jth energy storage system at time t; Nc is the number of energy storage systems; P load (t) is the total load of the power system at time t; The load balancing limitations are: ； Among them, P load工 (t) is the predicted power of the industrial load at time t; P load居 (t) is the predicted power of the life load at time t; P load (t) is the total load of the power system at time t; The grid power limit: ； Among them, P n (t) is the power supply of the power grid at time t; Minimum power constraints for supplying power to the grid; Maximum power constraints for supplying power to the grid; The photovoltaic power generation power limit is: ； Among them, P p.i (t) is the power generation of the ith photovoltaic power generation group at time t; is the minimum power constraint of the i-th photovoltaic power generation group; is the maximum power constraint of the i-th photovoltaic power generation group; The energy storage power supply power limit is: ； in, is the power supply power of the j-th energy storage system at time t; is the energy storage capacity of the j-th energy storage system at time t; is the energy storage capacity of the j-th energy storage system at time t-1; is the maximum capacity of the j-th energy storage system; The ramp rate limits are: ； Among them, Down n Up is the limit of the downward climbing rate of the power grid; n P is the upward climbing rate limit of the power grid; n (t) is the power supply of the power grid at time t; P n (t-1) is the power supply of the power grid at time t-1; Down p.i is the downward climbing rate limit of the ith photovoltaic power generation group; Up p.i is the upward climbing rate limit of the ith photovoltaic power generation group; P p.i (t)、P p.i (t-1) is the power generation of the ith photovoltaic power generation group at time t and time t-1; Down c.j Up is the downward climbing rate limit of the j-th energy storage system; c.j is the upward climbing rate limit of the j-th energy storage system; P c.j (t)、P c.j (t-1) is the power supply of the jth energy storage system at time t and time t-1; Conversion restrictions: ； ； in, is the predicted photovoltaic power of the i-th photovoltaic power generation group at time t; is the theoretical output power of the j-th energy storage system at time t; , is the conversion rate parameter.

4. The method for load dispatching of a park grid-connected operation according to claim 1, characterized in that: The set population exchange strategy includes: Step A1, determine whether it is the first exchange: If yes, proceed to step A2; If not, proceed to step A3; Step A2, at the first preset number of evolutionary generations, the results of the first N individuals, the middle N individuals and the last N individuals of the fitness value are respectively exchanged, where N is a positive integer; Step A3, at the second preset number of evolutionary generations, respectively take the results of the first M individuals and the last M individuals of the fitness value and exchange them, where M is a positive integer.

5. The method for load dispatching of park grid-connected operation according to claim 4, characterized in that: In step A2, the value of N increases as the current evolutionary generation increases.

6. The method for load dispatching of park grid-connected operation according to claim 4, characterized in that: In step A2, the value of N is determined using the following formula: ； Among them, z is the current evolutionary generation; S represents the population size; k represents the control coefficient; z represents the current evolutionary generation; represents the maximum number of evolutionary generations.

7. A park grid-connected load dispatching system, characterized in that: The method for dispatching loads in a park grid-connected operation as claimed in any one of claims 1 to 6 comprises: An acquisition unit, used for acquiring industrial load forecast power, domestic load forecast power and photovoltaic forecast power corresponding to a scheduling time period by using a preset forecast model; A model building unit, used to build a load dispatching model in a grid-connected operation state, wherein the load dispatching model includes a first dispatching target and a preset restriction condition; An optimization unit is used to generate a second scheduling target to update the objective function of the load scheduling model based on the electricity price information of the industrial load and the domestic load in the park during the peak and valley time periods and the power disconnection tolerance conditions, and configure corresponding time compensation values ​​for the industrial load forecast power and the domestic load forecast power respectively, and use the particle swarm algorithm and the firefly algorithm and the set population exchange strategy to solve the global optimal solution of the load scheduling model; The dispatching unit is used to reasonably configure the exchange energy between the photovoltaic power generation group, energy storage system, power consumption system and external power grid in the park according to the global optimal solution, so as to realize the load dispatching of the grid-connected operation system.

Images