A distributed photovoltaic group cooperative scheduling method and system

Abstract

The application discloses a kind of distributed photovoltaic group collaborative scheduling method and system, it is divided into multiple table areas according to the geographical position of roof photovoltaic distribution, and table area intelligent fusion terminal is deployed in table area control end, and the edge autonomy is realized based on the active-reactive joint sensitivity scheduling algorithm preset based on table area intelligent fusion terminal, the upper limit of dynamic apparent power of distribution transformer is replaced by calculating distribution transformer, and traditional static capacity constraint is adopted, and multi-layer Pareto optimization strategy is used, under the double safety constraints of guaranteeing voltage qualification and distribution transformer thermal stability, the active power consumption of photovoltaic is maximized, and photovoltaic light loss is significantly reduced.Simultaneously set cloud platform in collaborative scheduling center, only receive the emergency scheduling request sent by table area to realize table area level coordination scheduling, avoid the calculation power bottleneck when centralized scheduling center is dispatched to mass roof photovoltaic node, realize millisecond level on-site decision, minute level regional coordination.

Description

Technical Field

[0001] This invention relates to the field of collaborative scheduling of distributed photovoltaic (PV) clusters, and in particular to a method and system for collaborative scheduling of large-area distributed rooftop PV clusters. Background Technology

[0002] With the development of distributed photovoltaic (PV) systems, a large number of rooftop PV systems are being connected to low-voltage distribution networks in a distributed manner. Unlike traditional centralized PV power plants, rooftop PV systems are characterized by "numerous points and wide coverage, small capacity at each point, dispersed access locations, and strong output fluctuations," posing a severe challenge to the safe and stable operation of the distribution network.

[0003] In the grid-connected dispatch of a large number of rooftop photovoltaic systems, voltage overrun and reverse overload are the two most prominent operational risks. Voltage overrun refers to the period of peak photovoltaic power generation at midday when the load on low-voltage distribution areas is low. The reverse power injection causes a significant rise in the voltage at the end of the distribution area, which may exceed the voltage limit stipulated by national standards, affecting the safety of user equipment and triggering grid protection. Reverse overload refers to the backflow of large-scale photovoltaic power to the upper-level grid through the low-voltage side of the distribution transformer. When the reverse power exceeds the rated capacity of the transformer, it will cause the transformer windings to overheat, the insulation to age rapidly, and in severe cases, the equipment to be damaged.

[0004] Existing solutions for voltage overrun and reverse overload problems include installing anti-reverse current protection devices and setting independent active or reactive power regulation strategies. However, for voltage overrun, reactive power absorption needs to be increased, while for reverse overload, the total apparent power of distribution transformers needs to be limited. Furthermore, designs incorporating both solutions must consider the physical coupling conflicts between them. Current solutions treat the two problems as independent logics, lacking a unified, collaborative optimization approach. This often leads to scheduling results that are inconsistent: either limiting active power absorption for safety, or leaving excessive safety margins for distribution transformers, impacting generation efficiency.

[0005] Therefore, how to design a unified collaborative optimization governance method or system that, while ensuring that the voltage of the distribution transformer does not exceed the limit, generates as much active power as possible and absorbs reactive power to reduce voltage drop, so as to achieve the best balance between the safety margin of the distribution transformer and the photovoltaic absorption rate, and maximize the improvement of photovoltaic efficiency, is a technical problem that urgently needs to be solved in the existing technology. Summary of the Invention

[0006] The purpose of this invention is to overcome the shortcomings of existing technologies and propose a distributed photovoltaic (PV) cluster collaborative scheduling method and system. This method divides rooftop PV distribution into multiple zones based on their geographical location. Each zone has a smart fusion terminal deployed at its control point. Edge autonomy is achieved based on a pre-set active-reactive power joint sensitivity scheduling algorithm within these smart fusion terminals. The method replaces traditional static capacity constraints by calculating the dynamic apparent power limit of the distribution transformers and employs a multi-layer Pareto optimization strategy. Under the dual safety constraints of ensuring voltage compliance and distribution transformer thermal stability, it maximizes PV active power consumption and significantly reduces PV curtailment losses. Simultaneously, a cloud platform is set up in the collaborative scheduling center, receiving only emergency scheduling requests from the zones to achieve zone-level coordinated scheduling. This avoids the computational bottleneck of centralized scheduling centers when scheduling massive numbers of rooftop PV nodes, enabling millisecond-level local decision-making and minute-level regional coordination.

[0007] This invention provides a method for coordinated scheduling of distributed photovoltaic (PV) clusters, the method comprising:

[0008] Multiple distribution zones are set up according to the geographical location of the rooftop photovoltaic modules. Each distribution zone contains multiple rooftop photovoltaic modules, and the multiple rooftop photovoltaic modules are electrically connected to the power distribution network via the distribution zone distribution transformer.

[0009] Each distribution transformer area is equipped with a distribution transformer area control terminal, which is deployed with a distribution transformer area intelligent fusion terminal; the distribution transformer area intelligent fusion terminal is based on the collected three-phase voltage of the distribution transformer and the calculated thermal margin Q of the distribution transformer. m Determine whether both meet the operating conditions; if both meet the operating conditions, the intelligent converged terminal of the distribution area performs autonomous operation of the distribution area according to the preset hierarchical scheduling rules.

[0010] If the intelligent integrated terminal of the distribution area determines that either of the two conditions does not meet the operating conditions, the intelligent integrated terminal of the distribution area sends an emergency dispatch request to the collaborative dispatch center. The cloud platform of the collaborative dispatch center, based on the short-term load forecast data of each distribution area, the current actual total output power of the distribution area, and the adjustable capacity of energy storage / flexible load, issues dispatch strategy parameters to each intelligent integrated terminal of the distribution area to carry out distribution area-level coordinated dispatch.

[0011] Preferably, the intelligent fusion terminal for the distribution area is based on the collected three-phase voltage of the distribution transformer and the calculated thermal margin Q of the distribution transformer. m Determining whether both meet the operating conditions specifically includes:

[0012] The collected three-phase voltages of the distribution transformer show that none of the phase voltages exceeds the voltage warning threshold; and the calculated thermal margin Q of the distribution transformer is also included. m It exceeds the heat margin warning threshold.

[0013] Preferably, the intelligent converged terminal of the distribution area performs distribution area autonomy according to a preset hierarchical scheduling rule, specifically including:

[0014] The intelligent fusion terminal of the distribution area uses the distribution transformer thermal stability dynamic evaluation module to calculate the short-time dynamic apparent power limit S at the current moment. l and S l The active-reactive power is transmitted to the active-reactive power joint sensitivity scheduling module; the online sensitivity identification module calculates the active-reactive power sensitivity matrix of the distribution area and transmits it to the active-reactive power joint sensitivity scheduling module; the active-reactive power joint sensitivity scheduling module determines the active-reactive power based on the received short-time dynamic apparent power limit S. l The system comprehensively judges the active-reactive sensitivity matrix, the three-phase voltage of the distribution transformer, and the current actual total output power of the distribution area, and calls the corresponding scheduling layer of the hierarchical scheduling strategy for autonomous scheduling based on the comprehensive judgment result.

[0015] Preferably, the intelligent fusion terminal of the distribution area uses a distribution transformer thermal stability dynamic evaluation module to calculate the short-time dynamic apparent power limit S at the current moment. l Specifically, it includes:

[0016] The intelligent fusion terminal for the distribution area is based on the ambient temperature of the distribution transformer at the current moment. Calculate the hot spot temperature of the distribution transformer winding. :

[0017] ;

[0018] in, The current ambient temperature. For the current top oil temperature rise, The temperature rise of the top layer oil due to the hot spot of the winding;

[0019] Under steady-state conditions, the steady-state equation of a distribution transformer can be expressed as:

[0020] ;

[0021] ;

[0022] in, This refers to the temperature rise of the top oil under rated load. K is the current load rate, I S I represents the current current of the distribution transformer. n R is the rated current of the distribution transformer; R is the ratio of rated load loss to no-load loss; x is the oil index of the oil-immersed transformer. y represents the hotspot gradient; y represents the winding index.

[0023] Obtain the upper limit of the allowable winding hot spot temperature in the distribution transformer design. The maximum allowable temperature rise of the winding hot spot at the current moment is... for:

[0024] ;

[0025] Based on steady-state conditions, the temperature rise of the top layer oil due to the winding hot spot. By reverse-engineering the calculation formula, we can obtain the maximum allowable short-term load rate K at the current moment. a :

[0026] ;

[0027] The short-time dynamic apparent power limit S l for: Among them, S N This refers to the rated capacity of the distribution transformer.

[0028] in, ;K al The maximum short-term load rate allowed for actual scheduling at the current moment. This is the safety margin coefficient.

[0029] Preferably, the intelligent converged terminal for the distribution area uses an online sensitivity identification module to calculate the active-reactive power sensitivity matrix of the distribution area, specifically including:

[0030] The online sensitivity identification module calculates the voltage-active power sensitivity, voltage-reactive power sensitivity, apparent power-active power sensitivity, and apparent power-reactive power sensitivity of each inverter based on the collected power values ​​of each inverter within the transformer area and the transformer area's line topology.

[0031] The sensitivity parameters of the i-th inverter include voltage-active power sensitivity. The effect of the active power change of the i-th inverter on the terminal voltage U of the transformer area is represented. end The impact;

[0032] Voltage-Reactive Power Sensitivity , characterizing the effect of reactive power change of the i-th inverter on the terminal voltage U of the transformer area. end The impact;

[0033] Apparent power - active sensitivity , characterizing the effect of the active power change of the i-th inverter on the apparent power output of the distribution transformer;

[0034] Apparent power-reactive sensitivity , characterizing the impact of reactive power variation of the i-th inverter on the apparent power output of the distribution transformer;

[0035] The active-reactive power sensitivity matrix of the transformer area is obtained by weighted averaging the sensitivity parameters of each inverter within the transformer area. , , , ];in Indicates the overall voltage-active power sensitivity of the distribution area; Indicates the overall voltage-reactive power sensitivity of the transformer area; The overall apparent power-active power sensitivity of the transformer area; The apparent power-reactive power sensitivity of the entire distribution area.

[0036] Preferably, the active-reactive power joint sensitivity scheduling module performs a comprehensive judgment and, based on the comprehensive judgment result, calls the scheduling layer corresponding to the hierarchical scheduling strategy for autonomous scheduling, specifically including:

[0037] The active-reactive power joint sensitivity scheduling module will collect the three-phase voltage U at the outlet of the distribution transformer. a U b U c and the voltage over-limit threshold U respectively t Comparison and judgment; and acquisition of the current total active power P of the transformer area. t The current total reactive power Q in the Hetai area t Calculate the thermal margin Q of the distribution transformer. m and actual apparent power S a :

[0038] ;

[0039] ;

[0040] First-level scheduling rule: When the collected three-phase voltage U a U b U c All are less than the voltage over-limit threshold U t And the actual apparent power S a Less than the short-time dynamic apparent power limit S l If so, it will not interfere with the automatic operation of the equipment in the distribution area, and will continue to monitor and make judgments;

[0041] Second-level scheduling rule: When the collected three-phase voltage U a U b U c If any phase voltage exceeds the voltage over-limit threshold U t And the thermal margin of the distribution transformer In this case, only the reactive power absorption strategy is activated, without reducing the active power output of each inverter, thus maintaining full active power generation; Q nLet Q be the total reactive power to be absorbed; n The reactive power scheduling command is decomposed and converted into the reactive power scheduling command of each inverter and sent to each inverter for scheduling.

[0042] in ;

[0043] ; U represents the absolute value of the overall voltage-reactive power sensitivity of the transformer area; max For voltage limits, U d This is the voltage control dead zone value;

[0044] Third-level scheduling rule: When the collected three-phase voltage U a U b U c If any phase voltage exceeds the voltage over-limit threshold U t And the thermal margin of the distribution transformer A Pareto optimization model is constructed with the objective of maximizing active power consumption while satisfying the dual hard constraints of voltage not exceeding limits and thermal stability. The minimum active power reduction is then obtained by jointly searching for the optimal solution. and the corresponding reactive power absorption Find the minimum active power reduction. and the corresponding reactive power absorption Afterwards, the active power reduction and reactive power absorption of each inverter are calculated, and the corresponding power adjustment commands are sent to each inverter for adjustment and scheduling.

[0045] Fourth-level scheduling rule: When the collected three-phase voltage U a U b U c All are less than the voltage over-limit threshold U t And the actual apparent power S a Greater than or equal to the short-time dynamic apparent power limit S l Then, the minimum active power reduction can be found by simply reducing active power output to restore thermal stability. The active power of each inverter is converted into a scheduling command and sent to each inverter for scheduling.

[0046] Preferably, the second-level scheduling rule only activates the reactive power absorption strategy, specifically including:

[0047] The reactive power Q absorbed by the i-th inverter n,i for:

[0048] ; where Q i,a Let be the reactive power absorption capacity currently available for the i-th inverter, and j be the total number of inverters in the area.

[0049] Among them, S i,r Let P be the rated apparent power of the i-th inverter. i,current This represents the active power of the current i-th inverter.

[0050] Preferably, in the third-layer scheduling rules, the minimum active power reduction is obtained through joint optimization. and the corresponding reactive power absorption Specifically, it also includes:

[0051] The objective function is to minimize the amount of active power reduction. and the corresponding reactive power absorption ;

[0052] The hard constraint is:

[0053] 1. Voltage recovery constraint: ;in This represents the absolute value of the overall voltage-active power sensitivity of the transformer substation. This represents the absolute value of the overall voltage-reactive power sensitivity of the transformer area.

[0054] 2. Dynamic thermal stability constraints: ;

[0055] 3. Variable non-negativity constraint: , ;

[0056] 4. Adjustment range constraints: ;where P min,r The minimum active power that must be reserved for the transformer substation area;

[0057] ,in, S represents the maximum allowable reactive power absorption of the i-th inverter; i,r Let P be the rated apparent power of the i-th inverter. i,current Let be the active power of the i-th inverter;

[0058] According to voltage recovery constraints and The functional relationship between the components transforms the objective function into a function of a unique variable. Solving for:

[0059] ;

[0060] Will Substituting these values ​​into the dynamic thermal stability constraints, the minimum active power reduction can be quickly solved using the bisection method or the golden section method. ;

[0061] Find the minimum active power reduction. and the corresponding reactive power absorption Subsequently, for the active power reduction of each inverter, a proportional reduction method is adopted:

[0062] ;in, The reduction in active power allocated to the i-th inverter, P i,current Let be the active power of the i-th inverter, and j be the total number of inverters in the area;

[0063] The reactive power absorption amount is allocated according to the proportion of the remaining reactive power capacity of each inverter. :

[0064] ;in, Assign the corresponding reactive power absorption to the i-th inverter, Q i,a Let be the reactive power absorption capacity currently available for the i-th inverter.

[0065] Preferably, in the fourth-level scheduling rule, the minimum active power reduction is found by reducing active power output only to restore thermal stability. Specifically, it also includes:

[0066] The dynamic thermal stability constraint is: ;

[0067] Find the minimum active power reduction. To reduce the active power of each inverter, a proportional reduction method is adopted:

[0068] ;in, The reduction in active power allocated to the i-th inverter, P i,current Let be the active power of the i-th inverter, and j be the total number of inverters in the area.

[0069] The present invention also provides a distributed photovoltaic cluster collaborative scheduling system, the collaborative scheduling system comprising: multiple transformer substations and a collaborative scheduling center;

[0070] Each of the multiple transformer substations contains multiple rooftop photovoltaic modules, which are electrically connected to the power distribution network via the substation distribution transformer. Each substation is equipped with a substation control terminal, which is equipped with a substation smart integration terminal. The substation smart integration terminal includes a data acquisition module, an active-reactive power joint sensitivity scheduling module, and an execution module.

[0071] Among them, the active-reactive power joint sensitivity scheduling module of the intelligent integrated terminal of the distribution area first uses the three-phase voltage of the distribution transformer collected by the data acquisition module and the calculated thermal margin Q of the distribution transformer. mDetermine whether both meet the operating conditions; if both meet the operating conditions, the intelligent converged terminal of the distribution area performs autonomous operation of the distribution area according to the preset hierarchical scheduling rules, and sends the corresponding autonomous operation instructions to each inverter in the execution module to perform autonomous scheduling; if either of them does not meet the operating conditions, an emergency scheduling request is sent to the collaborative scheduling center.

[0072] The cloud platform of the collaborative dispatch center, based on the short-term load forecast data of each distribution area, the current actual total output power of the distribution area, and the adjustable capacity of energy storage / flexible load, sends dispatch strategy parameters to each distribution area's intelligent fusion terminal when it receives an emergency dispatch request. Each distribution area's intelligent fusion terminal then controls the execution module to perform dispatch actions based on its own dispatch strategy parameters to conduct distribution area-level coordinated dispatch.

[0073] The beneficial effects of this invention are as follows: 1. It can evaluate and calculate the heat capacity limit of the distribution transformer in real time, and use the short-time overload capacity designed for the distribution transformer to calculate how much heat capacity and time margin is left before the "real overload danger", which replaces the traditional "one-size-fits-all" static capacity limit and improves the grid's ability to accept photovoltaics.

[0074] 2. Traditional governance methods involve activating reactive power absorption to reduce voltage when the voltage exceeds the limit. However, absorbing reactive power increases the apparent power flowing through the transformer, which exacerbates the risk of reverse overload. To prevent overload, active power is directly reduced, causing unnecessary curtailment of solar power. These two methods conflict with each other in unified scheduling. This invention proposes an active-reactive power joint sensitivity scheduling algorithm, which incorporates both voltage constraints and thermal stability constraints into the same optimization problem for unified solution. It accurately calculates the quantitative relationship between "how much reactive power needs to be absorbed for voltage reduction" and "how much distribution transformer capacity will be occupied by absorbing this reactive power," thereby finding the minimum active power reduction that simultaneously satisfies both constraints, fundamentally eliminating the scheduling conflict between the two.

[0075] 3. To address the issues of unreliable line models caused by the numerous and widely distributed rooftop photovoltaic points, inaccurate line parameters, and frequent changes in topology due to load variations, a data-driven sensitivity identification method is adopted. This method estimates the voltage-active power, voltage-reactive power, apparent power-active power, and apparent power-reactive power sensitivity matrices in real time from actual operating data. It does not rely on offline physical parameters, can adapt to topology changes and parameter drift, and the sensitivity matrices are periodically updated to ensure control accuracy.

[0076] 4. Use heat margin Q m As a core performance indicator, it directly represents "how much additional reactive power the transformer can withstand, assuming the current active power output remains unchanged." This indicator is related to the Q value required for voltage control. n It allows for direct comparison, making the scheduling decision logic simple and intuitive, and enabling real-time and accurate monitoring of overload;

[0077] 5. The cloud platform only works when the intelligent integrated terminal of the distribution area monitors the three-phase voltage and the thermal margin Q of the distribution transformer. m Only when entering the "alert zone" does the system send an emergency dispatch request to the cloud platform or activate the locally pre-set autonomous strategy, pushing core optimization computing down to the intelligent converged terminal in the distribution area, reducing the computing pressure on the cloud platform and improving the system response speed. Attached Figure Description

[0078] Figure 1 A flowchart illustrating the steps of a distributed photovoltaic (PV) cluster collaborative scheduling method;

[0079] Figure 2 This is a schematic diagram of the structure of a distributed photovoltaic cluster collaborative scheduling system; Detailed Implementation

[0080] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings, but the scope of protection of the present invention is not limited to the following description.

[0081] This invention provides a method for coordinated scheduling of distributed photovoltaic (PV) clusters, such as... Figure 1 As shown, the cooperative scheduling method includes:

[0082] Multiple distribution zones are set up according to the geographical location of the rooftop photovoltaic modules. Each distribution zone contains multiple rooftop photovoltaic modules, which are electrically connected to the power distribution network via distribution transformers in the distribution zones. Specifically, the rooftop photovoltaic modules include photovoltaic inverters, wherein multiple rooftop photovoltaic modules that are geographically close are divided into one distribution zone.

[0083] Each distribution transformer area is equipped with a distribution transformer area control terminal, which is deployed with a distribution transformer area intelligent fusion terminal; the distribution transformer area intelligent fusion terminal is based on the collected three-phase voltage of the distribution transformer and the calculated thermal margin Q of the distribution transformer. m Determine whether both meet the operating conditions; if both meet the operating conditions, the intelligent converged terminal of the distribution area performs autonomous operation of the distribution area according to the preset hierarchical scheduling rules.

[0084] If the intelligent integrated terminal of the distribution area determines that either of the two conditions does not meet the operating conditions, the intelligent integrated terminal of the distribution area sends an emergency dispatch request to the collaborative dispatch center. The cloud platform of the collaborative dispatch center, based on the short-term load forecast data of each distribution area, the current actual total output power of the distribution area, and the adjustable capacity of energy storage / flexible load, issues dispatch strategy parameters to each intelligent integrated terminal of the distribution area to carry out distribution area-level coordinated dispatch.

[0085] Preferably, the intelligent fusion terminal for the distribution area is based on the collected three-phase voltage of the distribution transformer and the calculated thermal margin Q of the distribution transformer. m Determining whether both meet the operating conditions specifically includes:

[0086] The collected three-phase voltages of the distribution transformer show that none of the phase voltages exceeds the voltage warning threshold; and the calculated thermal margin Q of the distribution transformer is also included. m It exceeds the heat margin warning threshold.

[0087] Preferably, the intelligent converged terminal of the distribution area performs distribution area autonomy according to a preset hierarchical scheduling rule, specifically including:

[0088] The intelligent fusion terminal of the distribution area uses the distribution transformer thermal stability dynamic evaluation module to calculate the short-time dynamic apparent power limit S at the current moment. l and S l The active-reactive power is transmitted to the active-reactive power joint sensitivity scheduling module; the online sensitivity identification module calculates the active-reactive power sensitivity matrix of the distribution area and transmits it to the active-reactive power joint sensitivity scheduling module; the active-reactive power joint sensitivity scheduling module determines the active-reactive power based on the received short-time dynamic apparent power limit S. l The system comprehensively judges the active-reactive sensitivity matrix, the three-phase voltage of the distribution transformer, and the current actual total output power of the distribution area, and calls the corresponding scheduling layer of the hierarchical scheduling strategy for autonomous scheduling based on the comprehensive judgment result.

[0089] Specifically, the intelligent fusion terminal of the distribution area uses the distribution transformer thermal stability dynamic evaluation module to calculate the short-time dynamic apparent power limit S at the current moment. l Specifically, it includes:

[0090] The intelligent fusion terminal for the distribution area is based on the ambient temperature of the distribution transformer at the current moment. Calculate the hot spot temperature of the distribution transformer winding. :

[0091] ;

[0092] in, The current ambient temperature. For the current top oil temperature rise, The temperature rise of the top layer oil due to the hot spot of the winding;

[0093] Under steady-state conditions, the steady-state equation of a distribution transformer can be expressed as:

[0094] ;

[0095] ;

[0096] in, The temperature rise of the top oil under rated load can be obtained from the transformer nameplate or through experiments; K is the current load rate, I S I represents the current current of the distribution transformer.n R is the rated current of the distribution transformer; R is the ratio of rated load loss to no-load loss, both of which can be obtained from the transformer nameplate; x is the oil index of the oil-immersed transformer. The hotspot gradient can be obtained from the transformer nameplate or through experiments; y is the winding index, which can be set according to the transformer type, and can be selected from 1.3 to 1.6.

[0097] Under steady-state conditions, within a short period of time, such as 30 minutes, the top oil temperature of a distribution transformer changes very little due to its large thermal inertia. Therefore, when evaluating and calculating the maximum allowable short-term load rate of the distribution transformer at the current moment, it can be assumed that the top oil temperature rise remains unchanged, and the maximum allowable short-term load rate can be calculated by back-calculating using only the margin of the winding temperature rise.

[0098] Obtain the upper limit of the allowable winding hot spot temperature in the distribution transformer design. The maximum allowable temperature rise of the winding hot spot at the current moment is... for:

[0099] ;

[0100] Based on steady-state conditions, the temperature rise of the top layer oil due to the winding hot spot. By reverse-engineering the calculation formula, we can obtain the maximum allowable short-term load rate K at the current moment. a :

[0101] ;

[0102] The short-time dynamic apparent power limit S l for: Among them, S N This refers to the rated capacity of the distribution transformer.

[0103] To prevent control oscillations caused by temperature fluctuations at the boundary, a safety margin factor is set. Upper limit used for scheduling constraints:

[0104] in, ;K al The maximum short-term load rate allowed for actual scheduling at the current moment. As the safety margin coefficient, it can be taken as... =0.95.

[0105] In a low-voltage distribution area, each rooftop photovoltaic module contains an inverter node. The impact of inverter active power regulation on the voltage at the end of the distribution area depends on the line impedance between the inverter's connection point and the end of the distribution area. The farther the inverter's connection point is from the distribution transformer (the closer it is to the end), the greater the impact of its power changes on the end voltage. Conversely, the closer the inverter's connection point is to the distribution transformer, the more directly it affects the apparent power at the transformer's output. Sensitivity parameters can be used to characterize the differences in the inverter's location within the topology.

[0106] Preferably, the intelligent converged terminal for the distribution area uses an online sensitivity identification module to calculate the active-reactive power sensitivity matrix of the distribution area, specifically including:

[0107] The online sensitivity identification module calculates the voltage-active power sensitivity, voltage-reactive power sensitivity, apparent power-active power sensitivity, and apparent power-reactive power sensitivity of each inverter based on the collected power values ​​of each inverter within the transformer area and the transformer area's line topology.

[0108] The sensitivity parameters of the i-th inverter include voltage-active power sensitivity. The effect of the active power change of the i-th inverter on the terminal voltage U of the transformer area is represented. end The impact;

[0109] Voltage-Reactive Power Sensitivity , characterizing the effect of reactive power change of the i-th inverter on the terminal voltage U of the transformer area. end The impact;

[0110] Apparent power - active sensitivity , characterizing the effect of the active power change of the i-th inverter on the apparent power output of the distribution transformer;

[0111] Apparent power-reactive sensitivity , characterizing the impact of reactive power variation of the i-th inverter on the apparent power output of the distribution transformer;

[0112] The active-reactive power sensitivity matrix of the transformer area is obtained by weighted averaging the sensitivity parameters of each inverter within the transformer area. , , , ];in This represents the overall voltage-active power sensitivity of the distribution area, which is the weighted average of the voltage-active power sensitivity of each inverter. This represents the overall voltage-reactive power sensitivity of the distribution area, which is the weighted average of the voltage-reactive power sensitivity of each inverter. The apparent power-active power sensitivity of the entire distribution area is denoted as , and the apparent power-active power sensitivity of each inverter is denoted as . is the overall apparent power-reactive power sensitivity of the transformer area, and is the weighted average of the apparent power-reactive power sensitivity of each inverter.

[0113] During the trial operation phase of the transformer substation, due to the limited historical operating data, an active disturbance observation method was used for online identification. After the substation has been operating for a period of time, the natural fluctuations in photovoltaic output and load can be used as "excitation signals" without actively injecting disturbances. Historical operating data recorded by the intelligent integrated terminals in the substation were collected to construct a dataset. Partial least squares regression was used to train the sensitivity matrix offline / online. To adapt to topology changes and parameter drift, a rolling update was performed every 15 minutes, or an update calculation was triggered upon topology change to ensure control accuracy.

[0114] Preferably, the active-reactive power joint sensitivity scheduling module performs a comprehensive judgment and, based on the comprehensive judgment result, calls the scheduling layer corresponding to the hierarchical scheduling strategy for autonomous scheduling, specifically including:

[0115] The active-reactive power joint sensitivity scheduling module will collect the three-phase voltage U at the outlet of the distribution transformer. a U b U c and the voltage over-limit threshold U respectively t Comparison and judgment;

[0116] Among them, the voltage limit U is set according to national standards. max The voltage is 235V. To avoid frequent triggering and exiting of control commands, a voltage control dead zone U is set. d Typically, 1-2V is taken, then the voltage over-limit threshold U t for:

[0117] ;

[0118] And to collect the current total active power P of the transformer area t The current total reactive power Q in the Hetai area t Calculate the thermal margin Q of the distribution transformer. m and actual apparent power S a :

[0119] ;

[0120] ;

[0121] The heat margin Q of the distribution transformer m This refers to the current active power P t Under unchanged conditions, the distribution transformer shall not exceed the short-time dynamic apparent power limit S. l Under the premise of [unspecified condition], how much additional reactive power can it withstand? t Q represents the current total active power of the transformer area.t This represents the current total reactive power of the transformer substation.

[0122] First-level scheduling rule: When the collected three-phase voltage U a U b U c All are less than the voltage over-limit threshold U t And the actual apparent power S a Less than the short-time dynamic apparent power limit S l If so, it will not interfere with the automatic operation of the equipment in the distribution area, and will continue to monitor and make judgments;

[0123] Second-level scheduling rule: When the collected three-phase voltage U a U b U c If any phase voltage exceeds the voltage over-limit threshold U t And the thermal margin of the distribution transformer In this case, only the reactive power absorption strategy is activated, without reducing the active power output of each inverter, thus maintaining full active power generation; Q n Let Q be the total reactive power to be absorbed; n The reactive power scheduling command is decomposed and converted into the reactive power scheduling command of each inverter and sent to each inverter for scheduling.

[0124] in ;

[0125] ; U represents the absolute value of the overall voltage-reactive power sensitivity of the transformer area; max For voltage limits, U d This is the voltage control dead zone value;

[0126] Preferably, the second-level scheduling rule only activates the reactive power absorption strategy, specifically including:

[0127] The reactive power Q absorbed by the i-th inverter n,i for:

[0128] ; where Q i,a Let be the reactive power absorption capacity currently available for the i-th inverter, and j be the total number of inverters in the area.

[0129] Among them, S i,r Let P be the rated apparent power of the i-th inverter. i,current This represents the active power of the current i-th inverter.

[0130] Specifically, in calculating the total absorbed reactive power Q n Subsequently, a thermal stability constraint check was performed, and the apparent power S was calculated.a ’ :

[0131] ;

[0132] If S a ’ Less than or equal to the short-time dynamic apparent power limit S l If the verification passes, the reactive power to be absorbed is allocated according to the proportion of the remaining reactive power capacity of each inverter. If the verification fails, the third-level scheduling is executed.

[0133] Third-level scheduling rule: When the collected three-phase voltage U a U b U c If any phase voltage exceeds the voltage over-limit threshold U t And the thermal margin of the distribution transformer A Pareto optimization model is constructed with the objective of maximizing active power consumption while satisfying the dual hard constraints of voltage not exceeding limits and thermal stability. The minimum active power reduction is then obtained by jointly searching for the optimal solution. and the corresponding reactive power absorption Find the minimum active power reduction. and the corresponding reactive power absorption Afterwards, the active power reduction and reactive power absorption of each inverter are calculated, and the corresponding power adjustment commands are sent to each inverter for adjustment and scheduling.

[0134] Preferably, in the third-layer scheduling rules, the minimum active power reduction is obtained through joint optimization. and the corresponding reactive power absorption Specifically, it also includes:

[0135] The objective function is to minimize the amount of active power reduction. and the corresponding reactive power absorption ;

[0136] The hard constraint is:

[0137] 1. Voltage recovery constraint: ;in This represents the absolute value of the overall voltage-active power sensitivity of the transformer substation. This represents the absolute value of the overall voltage-reactive power sensitivity of the transformer area.

[0138] 2. Dynamic thermal stability constraints: ;

[0139] 3. Variable non-negativity constraint: , ;

[0140] 4. Adjustment range constraints: ;where P min,r The minimum active power that must be reserved for the transformer substation area;

[0141] ,in, S represents the maximum allowable reactive power absorption of the i-th inverter; i,r Let P be the rated apparent power of the i-th inverter. i,current Let be the active power of the i-th inverter;

[0142] According to voltage recovery constraints and The functional relationship between the components transforms the objective function into a function of a unique variable. Solving for:

[0143] ;

[0144] Will Substituting these values ​​into the dynamic thermal stability constraints, the minimum active power reduction can be quickly solved using the bisection method or the golden section method. ;

[0145] Specifically, in another embodiment, to simplify the computation, apparent power-active power sensitivity can be utilized. Apparent power-reactive sensitivity The dynamic thermal stability constraints are simplified:

[0146] ;in The apparent power-active power sensitivity of the entire distribution area is denoted as , and the apparent power-active power sensitivity of each inverter is denoted as . The apparent power-reactive power sensitivity of the entire transformer area is denoted as , and the apparent power-reactive power sensitivity of each inverter is denoted as .

[0147] Will Substituting these constraints into the simplified dynamic thermal stability framework, the minimum active power reduction can be quickly solved using the bisection method or the golden section method. ;

[0148] Find the minimum active power reduction. and the corresponding reactive power absorption Subsequently, for the active power reduction of each inverter, a proportional reduction method is adopted:

[0149] ;in, The reduction in active power allocated to the i-th inverter, P i,current Let be the active power of the i-th inverter, and j be the total number of inverters in the area;

[0150] The reactive power absorption amount is allocated according to the proportion of the remaining reactive power capacity of each inverter. :

[0151] ;in, Assign the corresponding reactive power absorption to the i-th inverter, Q i,a Let be the reactive power absorption capacity currently available for the i-th inverter.

[0152] The intelligent converged terminal of the distribution area issues corresponding power adjustment commands to each inverter based on the calculated active power reduction and reactive power absorption of each inverter for adjustment.

[0153] When the three-phase voltage U is detected and collected a U b U c All recovered to within the voltage over-limit threshold, and the actual apparent power S a Less than the short-time dynamic apparent power limit S l If the time exceeds a preset time threshold, such as 5 minutes, the active power output of each inverter will be gradually restored, with the restoration rate limited to no more than 0.1P. N,i / min; where P N,i It is the maximum active power that the i-th inverter is allowed to continuously output under standard test conditions, which can be obtained from the nameplate of each inverter.

[0154] Fourth-level scheduling rule: When the collected three-phase voltage U a U b U c All are less than the voltage over-limit threshold U t And the actual apparent power S a Greater than or equal to the short-time dynamic apparent power limit S l Then, the minimum active power reduction can be found by simply reducing active power output to restore thermal stability. The active power of each inverter is converted into a scheduling command and sent to each inverter for scheduling.

[0155] Preferably, in the fourth-level scheduling rule, the minimum active power reduction is found by reducing active power output only to restore thermal stability. Specifically, it also includes:

[0156] Since the voltage is normal, there is no need to consider voltage constraints in this state. Therefore, the dynamic thermal stability constraint is set as follows: ;

[0157] Find the minimum active power reduction. To reduce the active power of each inverter, a proportional reduction method is adopted:

[0158] ;in, The reduction in active power allocated to the i-th inverter, Pi,current Let be the active power of the i-th inverter, and j be the total number of inverters in the area.

[0159] During the active power reduction process, each inverter verifies the reduced active power. If the inverter has reached its active power output limit, the remaining reduction is transferred to other inverters that still have a reduction margin. When the actual apparent power S... a Less than the short-time dynamic apparent power limit S l If the time exceeds a preset time threshold, such as 5 minutes, the active power output of each inverter will be gradually restored, with the restoration rate limited to no more than 0.1P. N,i / min; where P N,i It is the maximum active power that the i-th inverter is allowed to continuously output under standard test conditions, which can be obtained from the nameplate of each inverter.

[0160] Specifically, if the intelligent converged terminal of the distribution area determines that either of the two conditions does not meet the operating requirements, the intelligent converged terminal of the distribution area sends an emergency dispatch request to the collaborative dispatch center, which specifically includes:

[0161] The collected three-phase voltage U of the distribution transformer a U b U c Exceeding the voltage warning threshold, or the calculated thermal margin Q of the distribution transformer. m When the thermal margin warning threshold is less than the local preset rapid protection autonomous strategy, the active power output is directly reduced to a preset depth to ensure equipment safety. For example, the active power output is immediately reduced to 50% of the rated active power output, and an emergency dispatch request is sent to the cloud platform at the same time to request support and coordination strategy.

[0162] The voltage warning threshold is set at 1.1U. 标称相电压 The thermal margin Q of the distribution transformer m The alert threshold is -0.1 seconds. N’ S N’ The rated capacity S of the distribution transformer N The value.

[0163] In one embodiment, the intelligent fusion terminal of the distribution area also includes a load forecasting module, which performs short-term load forecasting based on the acquired meteorological data and sends it to the cloud platform. After receiving the emergency dispatch request, the cloud platform immediately performs regional coordination based on the received short-term load forecast data of each distribution area, the current actual total output power of the distribution area, and the adjustable capacity of energy storage / flexible load. It calculates the total active power and total reactive power of each distribution area, sends dispatching strategy parameters to the intelligent fusion terminal of each distribution area, and notifies adjacent distribution areas with reactive power margins to increase reactive power support or adjust the energy storage system for charging and discharging.

[0164] The intelligent converged terminal of the distribution area controls the execution modules within the area, such as inverters, energy storage, reactive power generators, flexible loads, and tie switches, to take action based on the corresponding dispatch strategy parameters received, so as to achieve coordinated dispatch at the distribution area level.

[0165] This invention also provides a distributed photovoltaic cluster collaborative scheduling system, such as... Figure 2 As shown, the collaborative scheduling system includes: multiple distribution areas and a collaborative scheduling center;

[0166] Each of the multiple transformer substations contains multiple rooftop photovoltaic modules, which are electrically connected to the power distribution network via the substation distribution transformer. Each substation is equipped with a substation control terminal, which is equipped with a substation smart integration terminal. The substation smart integration terminal includes a data acquisition module, an active-reactive power joint sensitivity scheduling module, and an execution module.

[0167] Among them, the active-reactive power joint sensitivity scheduling module of the intelligent integrated terminal of the distribution area first uses the three-phase voltage of the distribution transformer collected by the data acquisition module and the calculated thermal margin Q of the distribution transformer. m Determine whether both meet the operating conditions; if both meet the operating conditions, the intelligent converged terminal of the distribution area performs autonomous operation of the distribution area according to the preset hierarchical scheduling rules, and sends the corresponding autonomous operation instructions to each inverter in the execution module to perform autonomous scheduling; if either of them does not meet the operating conditions, an emergency scheduling request is sent to the collaborative scheduling center.

[0168] The cloud platform of the collaborative dispatch center, based on the short-term load forecast data of each distribution area, the current actual total output power of the distribution area, and the adjustable capacity of energy storage / flexible load, sends dispatch strategy parameters to each distribution area's intelligent fusion terminal when it receives an emergency dispatch request. Each distribution area's intelligent fusion terminal then controls the execution module to perform dispatch actions based on its own dispatch strategy parameters to conduct distribution area-level coordinated dispatch.

[0169] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, ROM, RAM, etc.

[0170] The above description discloses only preferred embodiments of the present invention and should not be construed as limiting the scope of the present invention. Therefore, equivalent variations made in accordance with the claims of the present invention are still within the scope of the present invention.

Claims

1. A method for coordinated scheduling of distributed photovoltaic (PV) clusters, characterized in that: The cooperative scheduling method includes: Multiple distribution zones are set up according to the geographical location of the rooftop photovoltaic modules. Each distribution zone contains multiple rooftop photovoltaic modules, and the multiple rooftop photovoltaic modules are electrically connected to the power distribution network via the distribution zone distribution transformer. Each distribution transformer area is equipped with a distribution transformer area control terminal, which is deployed with a distribution transformer area intelligent fusion terminal; the distribution transformer area intelligent fusion terminal is based on the collected three-phase voltage of the distribution transformer and the calculated thermal margin Q of the distribution transformer. m Determine whether both meet the operating conditions; if both meet the operating conditions, the intelligent converged terminal of the distribution area performs autonomous operation of the distribution area according to the preset hierarchical scheduling rules. If the intelligent integrated terminal of the distribution area determines that either of the two conditions does not meet the operating conditions, the intelligent integrated terminal of the distribution area sends an emergency dispatch request to the collaborative dispatch center. The cloud platform of the collaborative dispatch center, based on the short-term load forecast data of each distribution area, the current actual total output power of the distribution area, and the adjustable capacity of energy storage / flexible load, issues dispatch strategy parameters to each intelligent integrated terminal of the distribution area to carry out distribution area-level coordinated dispatch.

2. The distributed photovoltaic cluster collaborative scheduling method according to claim 1, characterized in that: The intelligent fusion terminal for the distribution area is based on the collected three-phase voltage of the distribution transformer and the calculated thermal margin Q of the distribution transformer. m Determining whether both meet the operating conditions specifically includes: The collected three-phase voltages of the distribution transformer show that none of the phase voltages exceeds the voltage warning threshold; and the calculated thermal margin Q of the distribution transformer is also included. m It exceeds the heat margin warning threshold.

3. The distributed photovoltaic cluster collaborative scheduling method according to claim 1, characterized in that: The intelligent converged terminal in the distribution area performs autonomous operation according to preset hierarchical scheduling rules, specifically including: The intelligent fusion terminal of the distribution area uses the distribution transformer thermal stability dynamic evaluation module to calculate the short-time dynamic apparent power limit S at the current moment. l and S l The active-reactive power is transmitted to the active-reactive power joint sensitivity scheduling module; the online sensitivity identification module calculates the active-reactive power sensitivity matrix of the distribution area and transmits it to the active-reactive power joint sensitivity scheduling module; the active-reactive power joint sensitivity scheduling module determines the active-reactive power based on the received short-time dynamic apparent power limit S. l The system comprehensively judges the active-reactive sensitivity matrix, the three-phase voltage of the distribution transformer, and the current actual total output power of the distribution area, and calls the corresponding scheduling layer of the hierarchical scheduling strategy for autonomous scheduling based on the comprehensive judgment result.

4. The distributed photovoltaic cluster collaborative scheduling method according to claim 3, characterized in that: The intelligent fusion terminal of the distribution area uses the distribution transformer thermal stability dynamic evaluation module to calculate the short-time dynamic apparent power limit S at the current moment. l Specifically, it includes: The intelligent fusion terminal for the distribution area is based on the ambient temperature of the distribution transformer at the current moment. Calculate the hot spot temperature of the distribution transformer winding. : ； in, The current ambient temperature. For the current top oil temperature rise, The temperature rise of the top layer oil due to the hot spot of the winding; Under steady-state conditions, the steady-state equation of a distribution transformer can be expressed as: ； ； in, This refers to the temperature rise of the top oil under rated load. K is the current load rate, I S I represents the current current of the distribution transformer. n R is the rated current of the distribution transformer; R is the ratio of rated load loss to no-load loss; x is the oil index of the oil-immersed transformer. y represents the hotspot gradient; y represents the winding index. Obtain the upper limit of the allowable winding hot spot temperature in the distribution transformer design. The maximum allowable temperature rise of the winding hot spot at the current moment is... for: ； Based on steady-state conditions, the temperature rise of the top layer oil due to the winding hot spot. By reverse-engineering the calculation formula, we can obtain the maximum allowable short-term load rate K at the current moment. a : ； The short-time dynamic apparent power limit S l for: Among them, S N This refers to the rated capacity of the distribution transformer. in, ;K al The maximum short-term load rate allowed for actual scheduling at the current moment. This is the safety margin coefficient.

5. The distributed photovoltaic cluster collaborative scheduling method according to claim 3, characterized in that: The intelligent fusion terminal for the distribution area uses an online sensitivity identification module to calculate the active-reactive power sensitivity matrix of the distribution area, specifically including: The online sensitivity identification module calculates the voltage-active power sensitivity, voltage-reactive power sensitivity, apparent power-active power sensitivity, and apparent power-reactive power sensitivity of each inverter based on the collected power values ​​of each inverter within the transformer area and the transformer area's line topology. The sensitivity parameters of the i-th inverter include voltage-active power sensitivity. The effect of the active power change of the i-th inverter on the terminal voltage U of the transformer area is represented. end The impact; Voltage-Reactive Power Sensitivity , characterizing the effect of reactive power change of the i-th inverter on the terminal voltage U of the transformer area. end The impact; Apparent power - active sensitivity , characterizing the effect of the active power change of the i-th inverter on the apparent power output of the distribution transformer; Apparent power-reactive sensitivity , characterizing the impact of reactive power variation of the i-th inverter on the apparent power output of the distribution transformer; The active-reactive power sensitivity matrix of the transformer area is obtained by weighted averaging the sensitivity parameters of each inverter within the transformer area. , , , ];in Indicates the overall voltage-active power sensitivity of the distribution area; Indicates the overall voltage-reactive power sensitivity of the transformer area; The overall apparent power-active power sensitivity of the transformer area; The apparent power-reactive power sensitivity of the entire distribution area.

6. The distributed photovoltaic cluster collaborative scheduling method according to claim 3, characterized in that: The active-reactive power joint sensitivity scheduling module performs a comprehensive judgment and, based on the comprehensive judgment result, calls the scheduling layer corresponding to the hierarchical scheduling strategy for autonomous scheduling, specifically including: The active-reactive power joint sensitivity scheduling module will collect the three-phase voltage U at the outlet of the distribution transformer. a U b U c and the voltage over-limit threshold U respectively t Comparison and judgment; and acquisition of the current total active power P of the transformer area. t The current total reactive power Q in the Hetai area t Calculate the thermal margin Q of the distribution transformer. m and actual apparent power S a : ； ； First-level scheduling rule: When the collected three-phase voltage U a U b U c All are less than the voltage over-limit threshold U t And the actual apparent power S a Less than the short-time dynamic apparent power limit S l If so, it will not interfere with the automatic operation of the equipment in the distribution area, and will continue to monitor and make judgments; Second-level scheduling rule: When the collected three-phase voltage U a U b U c If any phase voltage exceeds the voltage over-limit threshold U t And the thermal margin of the distribution transformer In this case, only the reactive power absorption strategy is activated, without reducing the active power output of each inverter, thus maintaining full active power generation; Q n Let Q be the total reactive power to be absorbed; n The reactive power scheduling command is decomposed and converted into the reactive power scheduling command of each inverter and sent to each inverter for scheduling. in ; ; U represents the absolute value of the overall voltage-reactive power sensitivity of the transformer area; max For voltage limits, U d This is the voltage control dead zone value; Third-level scheduling rule: When the collected three-phase voltage U a U b U c If any phase voltage exceeds the voltage over-limit threshold U t And the thermal margin of the distribution transformer A Pareto optimization model is constructed with the objective of "maximizing active power consumption while satisfying the dual hard constraints of voltage not exceeding limits and thermal stability". The minimum active power reduction is obtained by jointly searching for the optimal solution. and the corresponding reactive power absorption Find the minimum active power reduction. and the corresponding reactive power absorption Afterwards, the active power reduction and reactive power absorption of each inverter are calculated, and the corresponding power adjustment commands are sent to each inverter for adjustment and scheduling. Fourth-level scheduling rule: When the collected three-phase voltage U a U b U c All are less than the voltage over-limit threshold U t And the actual apparent power S a Greater than or equal to the short-time dynamic apparent power limit S l Then, the minimum active power reduction can be found by simply reducing active power output to restore thermal stability. The active power of each inverter is converted into a scheduling command and sent to each inverter for scheduling.

7. A distributed photovoltaic cluster collaborative scheduling method according to claim 6, characterized in that: The second-level scheduling rules only activate the reactive power absorption strategy, specifically including: The reactive power Q absorbed by the i-th inverter n,i for: ; where Q i,a Let be the reactive power absorption capacity currently available for the i-th inverter, and j be the total number of inverters in the area. Among them, S i,r Let P be the rated apparent power of the i-th inverter. i,current This represents the active power of the current i-th inverter.

8. A distributed photovoltaic cluster collaborative scheduling method according to claim 6, characterized in that: In the third-level scheduling rules, the minimum active power reduction is obtained through joint optimization. and the corresponding reactive power absorption Specifically, it also includes: The objective function is to minimize the active power reduction. and the corresponding reactive power absorption ; The hard constraint is:

1. Voltage recovery constraint: ;in This represents the absolute value of the overall voltage-active power sensitivity of the transformer substation. This represents the absolute value of the overall voltage-reactive power sensitivity of the transformer area.

2. Dynamic thermal stability constraints: ; 3. Variable non-negativity constraint: , ; 4. Adjustment range constraints: ;where P min,r The minimum active power that must be reserved for the transformer substation area; ,in, S represents the maximum allowable reactive power absorption of the i-th inverter; i,r Let P be the rated apparent power of the i-th inverter. i,current Let be the active power of the i-th inverter; According to voltage recovery constraints and The functional relationship between the components transforms the objective function into a function of a unique variable. Solving for: ; Will Substituting these values ​​into the dynamic thermal stability constraints, the minimum active power reduction can be quickly solved using the bisection method or the golden section method. ; Find the minimum active power reduction. and the corresponding reactive power absorption Subsequently, for the active power reduction of each inverter, a proportional reduction method is adopted: ;in, The reduction in active power allocated to the i-th inverter, P i,current Let be the active power of the i-th inverter, and j be the total number of inverters in the area; The reactive power absorption amount is allocated according to the proportion of the remaining reactive power capacity of each inverter. : ;in, Assign the corresponding reactive power absorption to the i-th inverter, Q i,a Let be the reactive power absorption capacity currently available for the i-th inverter.

9. A distributed photovoltaic cluster collaborative scheduling method according to claim 6, characterized in that: In the fourth-level scheduling rule, thermal stability is restored only by reducing active power output, and the minimum active power reduction is calculated. Specifically, it also includes: The dynamic thermal stability constraint is: ; Find the minimum active power reduction. To reduce the active power of each inverter, a proportional reduction method is adopted: ;in, The reduction in active power allocated to the i-th inverter, P i,current Let be the active power of the i-th inverter, and j be the total number of inverters in the area.

10. A distributed photovoltaic cluster collaborative scheduling system, characterized in that: The collaborative scheduling system includes: multiple distribution areas and a collaborative scheduling center; Each of the multiple transformer substations contains multiple rooftop photovoltaic modules, which are electrically connected to the power distribution network via the substation distribution transformer. Each substation is equipped with a substation control terminal, which is equipped with a substation smart integration terminal. The substation smart integration terminal includes a data acquisition module, an active-reactive power joint sensitivity scheduling module, and an execution module. Among them, the active-reactive power joint sensitivity scheduling module of the intelligent integrated terminal of the distribution area first uses the three-phase voltage of the distribution transformer collected by the data acquisition module and the calculated thermal margin Q of the distribution transformer. m Determine whether both meet the operating conditions; if both meet the operating conditions, the intelligent converged terminal of the distribution area performs autonomous operation of the distribution area according to the preset hierarchical scheduling rules, and sends the corresponding autonomous operation instructions to each inverter in the execution module to perform autonomous scheduling; if either of them does not meet the operating conditions, an emergency scheduling request is sent to the collaborative scheduling center. The cloud platform of the collaborative dispatch center, based on the short-term load forecast data of each distribution area, the current actual total output power of the distribution area, and the adjustable capacity of energy storage / flexible load, sends dispatch strategy parameters to each distribution area's intelligent fusion terminal when it receives an emergency dispatch request. Each distribution area's intelligent fusion terminal then controls the execution module to perform dispatch actions based on its own dispatch strategy parameters to conduct distribution area-level coordinated dispatch.

Images