Application of simulated annealing algorithm in multi-objective cooperative scheduling of load and storage of source network for load side of new power system

Abstract

To improve the adaptability of grid load collaborative scheduling, a multi-objective collaborative scheduling method based on a simulated annealing algorithm for the load storage of grid loads on the load side of a new power system is proposed. Local bus transmission technology is adopted to collect the dynamic parameters of energy network load energy storage on the load side of the new power system. The collected load dynamic parameters are fused with energy distribution state parameters to extract the state characteristics of energy network load storage. The simulated annealing algorithm is adopted to realize the load characteristics fusion and adaptive scheduling processing of energy network on the load side of the power system, and the spectral characteristics of the load dynamic parameters are extracted. The dynamic scheduling method of simulated annealing is used to realize the multi-objective optimization of dynamic load of energy network. Based on the co-optimization results of simulated annealing, the optimization application of the simulated annealing algorithm in the multi-objective co-scheduling of loads and energy storage in a new power system is realized. The experimental results show that after 400 iterations, the control convergence accuracy of the proposed method reaches 0.980, which is significantly better than that of the comparison method, and performs well in terms of scheduling efficiency improvement, load scheduling stability, scheduling time and energy waste ratio, proving that the method has good multi-objective integration and strong optimization ability in the scheduling process, and improves the load balanced scheduling and adaptive control ability of the power system.

Introduction

With the development of energy-saving control technology, research on the energy network load energy storage calculation of the new power system load side has received attention. Through energy network load energy storage calculations of the new power system load side, the energy-saving control ability is improved [1]. According to the development scale of the energy network load energy storage technology of the new power system load side, effective multi-objective collaborative scheduling of the energy network load energy storage of the new power system load side is needed [2].With the development of energy-saving control technology, the study of new power system load-side grid load storage calculations has received increasing attention because it can improve the energy-saving control capability. According to the scale of the development of this technology, effective multi-objective cooperative scheduling is required [3]. A new power system is an energy infrastructure that integrates a high proportion of renewable energy sources advanced energy storage technologies, intelligent grid management systems, and efficient energy conversion and distribution mechanisms [4]. It is designed to meet the challenges of sustainable energy development, with the aim of achieving high energy efficiency, low carbon emissions, reliable power supply, and flexible grid operation[5,6,7].

Currently, many scholars have conducted research on multi-objective cooperative scheduling methods for load storage in load-side energy networks of new power systems. Xu Yan et al. [8] demand response and interconnection line interaction strategies are introduced to establish a two-tier two-stage dispatch optimization model. The active network loss of the upper distribution network is minimized, the first stage of the lower microgrid optimizes the comprehensive economic and environmental indicators, and the second stage minimizes the power imbalance. The model is simplified using a second-order cone and cohesion function to develop the NSGA-II algorithm. The MOSEK and Bonmin optimization toolkits are invoked to solve and validate the constructed distribution and microgrid case systems to analyze the sensitivity of various factors to the optimal scheduling results. After using this method for scheduling strategy, it can effectively achieve the coordinated optimization of multi-objective functions in the distribution microgrid while improving the power quality of the distribution network and the comprehensive benefits of the microgrid. Chen Fei [9] used a multi-source complementary hierarchical co-optimization (HCOS) scheduling method based on multi-objective genetic particle swarm optimization for preprocessing, storing and displaying power data. It can provide a more intuitive understanding of the specific operation of diverse power systems. It realizes a large overall output of wind-solar energy storage, reduces overall power generation costs, and improves the economy and security of the power system. By establishing relevant constraints, a power system can operate safely, stably, and efficiently. Li Linglin et al. [10] proposed an improved tunica group algorithm to solve the model, introducing Levy flight strategy to enhance the optimization ability of the algorithm and adding nonlinear convergence factors to improve the convergence ability of the algorithm; Using a power system consisting of six generator sets and fifteen generator sets as test cases, the optimization effect of the improved algorithm on load dispatch problems is verified The results show that this method can solve the power load economic dispatch problem of two testing systems faster and more effectively, with the best optimization effect, the strongest optimization ability, and better operational stability. Huang Wenqi et al. [11] proposed a prediction method that integrates variable selection and sparse Transformer model. Static and temporal variables are used as inputs to fully utilize the information enhancement effect of static variables in the global time range. Based on the gating mechanism, a variable decentralization component is designed to assign different weights to variables according to their correlation with the prediction results, and a dual layer encoding structure is designed for temporal feature extraction, sparse attention processing, and prediction of future time loads through multivariate inputs. Experiments based on real power load data show that the model can improve the accuracy and efficiency of medium—and long-term load forecasting.

This paper introduces a new technology for calculating energy network load energy storage based on continuous energy saving, aimed at improving energy efficiency and reducing waste through optimization algorithms and control strategies. For load-side energy management in new power systems, multi-objective collaborative scheduling is crucial due to the dynamic and complex nature of energy demands and various loads. This approach effectively manages the intermittency and variability of renewable energy by coordinating energy storage devices, flexible loads, and demand response programs to ensure a stable power supply and better alignment between energy generation and consumption. Multi-objective collaborative scheduling is vital for efficient and sustainable load-side energy management. Initially, local bus transmission technology collects energy storage parameters from the load side of the new power system, which are then analyzed using nonlinear time series. A spectral parameter feature extraction method detects load energy storage characteristics. Clustering analysis of these characteristics is performed using a continuous energy saving parameter feature analysis method, leading to the construction of a feature reconstruction model for time series energy storage of grid loads. This enables accurate energy storage calculations for the load side of the new power system. Simulation tests demonstrate the superior performance of this method in enhancing the calculation accuracy of load-side energy storage.

The following are the specific novelties of this article:

• 1. Propose a multi-objective collaborative scheduling method based on simulated annealing algorithm: This paper proposes for the first time a solution based on a simulated annealing algorithm for the multi-objective collaborative scheduling problem of load energy storage in the load side energy network of the new power system.

• 2. Integrating local bus transmission technology with energy distribution state parameters: In this study, local bus transmission technology is used to realize the real-time acquisition of dynamic parameters of load storage energy in the load side energy network of a new type of power system. These dynamic parameters are fused using the energy distribution state parameters to extract the state characteristics of load storage in the energy network.

• 3. Realize load characteristic fusion and adaptive scheduling processing:

By adopting the dynamic scheduling method, multi-objective cooperative optimization of load storage in load-side energy networks in new energy systems is realized, which significantly improves the adaptability of multi-objective cooperative scheduling and makes the scheduling process more flexible and efficient. By fusing the load characteristics, precise control of the dynamic loads is realized, which improves the overall operational efficiency and stability of the power system.

This paper is organized as follows: The first part introduces the research background, highlighting issues with current scheduling methods and proposing a multi-objective cooperative scheduling method based on the simulated annealing algorithm. The second part uses local bus transmission technology to collect load storage parameters from the energy network on the load side of the new power system, establishes an energy parameter analysis model and statistical time series fusion model, and fuses dynamic parameters with state parameters to extract state characteristic quantities. The third part establishes an analytical model to obtain the conditional probability density function and achieves dynamic load multi-objective optimization through simulated annealing dynamic scheduling. The fourth part compares traditional methods to demonstrate the advantages of this paper’s multi-objective co-scheduling method. The fifth part summarizes the advantages of the proposed method and verifies its superior performance with experimental results.

Time series based analysis of new power system load side source grid load storage

Collection of equilibrium parameters

A source network refers to a network of energy sources rather than a the traditional upstream network. It is a network that supplies energy to the load side. A new power system refers to one that adopts new technologies, functions, or operational concepts. This is not just about the planning of the power system, but also a system that develops with progress, such as improved energy storage capacity, better load management strategies, and enhanced adaptability through the techniques described in this article, such as using simulated annealing algorithms for multi-objective collaborative scheduling of loads and energy storage. A dynamic gain parameter analysis model of the electric unit load was constructed to construct the dynamic load adjustment model based on new energy consumption The Newton–Raphson method was used to construct a parameter analysis control model of load storage in the load side energy network of the new electric power system, and the sag control eigenvolume considering the frequency variable was established [10]. The dynamic parameter matching set of load storage energy of load side energy network of the new power company is obtained, and the parameter collection model of load storage energy of load side energy network of the new power grid is constructed. Local bus transmission technology is adopted to collect the energy storage parameters of the load side of a new power system [11]. Local bus technology refers to a technological approach used to collect the energy storage parameters of the new power system load side. This enables the transfer of data related to energy network load energy storage. This technology plays a crucial role in gathering the necessary information for further analysis and processing. It allows for the efficient acquisition of dynamic parameters, which are then fused with the energy distribution state parameters to extract valuable state characteristics of the energy network load energy storage. By facilitating data collection and parameter fusion, local bus technology contributes to the overall understanding and management of the energy network load energy storage system in the new power system.

Based on the results of the parameter acquisition and information fusion, the energy storage characteristics of the load side of the new power system were analyzed and calculated. An overall structural model of the load-side energy storage calculation of the new power system is obtained, as shown in Fig. 1.

Fig. 1

Overall structural model of energy network load energy storage calculation on load side of new power system

As shown in Fig. 1, the first step is to collect the load storage parameters of the load side energy network of the new power system using the bus transmission control protocol. Based on the analysis of the load side energy network load storage state parameters of the new power system, the voltage vector is successfully switched to vector No. 2. Multi-objective coordinated parameter tuning of load storage of load side energy network of new power company. The objectives of multi-objective scheduling include enhancing the adaptability of multi-objective collaborative scheduling of load storage in the load-side energy network of a new power system, making the scheduling process more flexible and efficient, and realizing multi-objective optimization of dynamic loads in the energy network to improve the overall operational efficiency and stability of the power system. Quantile regression, load point prediction and residual analysis are directly carried out on the load, and the time window interval $\Delta = T_{1} - T_{2}$ of energy network load energy storage collection is obtained. Then the energy network load energy storage time series on the load side of the new power system is expressed as:

$$X(t) = \left( {\begin{array}{*{20}c} {x_{1} (t)} \\ \vdots \\ {x_{m} (t)} \\ \end{array} } \right)$$

(1)

wherein, $t$ represents the time variable of the time series.

Using the method of forecasting model and user load probability forecasting analysis, the distribution of frequency and voltage parameters in the no-load time of load probability forecasting is obtained as follows:

$$H_{1} = \frac{M\cos \alpha }{{4\pi \mu_{0} R^{5} }}[2(2h)^{2} - a^{2} ]$$

(2)

$$H_{2} = \frac{M\sin \alpha }{{4\pi \mu_{0} R^{5} }}[ - (4h^{2} + a^{2} )]$$

(3)

$$H_{rc} = \frac{M\cos 2\theta }{{4\pi \mu_{0} R^{5} }}(2h^{2} - a^{2} )$$

(4)

wherein,$h$ represents dynamic magnetic density;$\mu_{0}$ represents the amplitude of positive sequence voltage; $M$ represents the energy network load storage parameters; $a$ represents the equilibrium parameter; $\alpha$, $\theta$ represent the set of dynamic characteristic distributions; $R$ represents the voltage parameter.

While $h = \frac{a}{\sqrt 2 }$, $H_{rc} = 0$. From the perspective of load forecasting techniques and load probability prediction, the dynamic feature distribution set of the load storage energy of load side grid is obtained [12]. Combined with the state parameter fusion, the dynamic magnetic density S of load storage in the load-side energy network of the new power system is obtained as $B_{sy}^{knee}$. The positive sequence voltage amplitude is introduced, and the equilibrium parameter $B_{r}$ of energy network load energy storage on the load side of new power system is obtained according to the generated load quantile prediction result [13]. A multilayer perceptron with a hidden layer distribution approach is used for dynamic coordination based on the collection results of load storage parameters of the grid load on the load side of the new power system [14].

Load energy storage state fusion

Energy network load storage refers to a load-related energy storage component or mechanism in an energy network. It differs from a typical energy storage device in that it is specifically integrated with the load side of the energy network and designed according to the characteristics and requirements of the load. Unlike active loads, which consume electrical energy, energy network load storage stores energy and releases it back into the network when needed. The net load is indirectly predicted, and according to the unsupervised BTM photovoltaic decomposition method, the probability distribution and parameters of reliability evaluation and prediction of energy network load storage on the load side of power system are obtained as follows:

$$K(t) = \{ X_{d + 1} ,X_{d + 2} , \cdot \cdot \cdot X_{d + m} \}$$

(5)

Based on the series and parallel joint analytical methods, it is obtained that the fuzzy iteration number $V = [V_{1} ,V_{2} , \cdot \cdot \cdot ,V_{m} ] \in R^{m \times m}$ of the energy network load energy storage state space on the load side of the new power system is the minimum. When $V \in R^{m \times m}$, energy network load storage on the load side of the new power system has a minimum value, the net load and actual net load are used as the target measurements. The results of the joint probability density characterization of the load side of the new power system are as follows:

$$y(t) = \frac{1}{\pi }P\int {\left( {\begin{array}{*{20}c} {x_{1} (t)} \\ \vdots \\ {x_{m} (t)} \\ \end{array} } \right)} d\tau = x(t)\sqrt {\frac{1}{{\pi f_{sw} \mu_{0} \delta }}}$$

(6)

Using an unsupervised data-driven method, a new energy parameter analysis model of the power system load side is established, and the statistical time series fusion model of the joint characteristic analysis of load and residual is obtained as follows:

$$\begin{gathered} G\left( f \right) = \frac{{T_{B} }}{{\left( {\rho \frac{N \cdot MLT}{{\pi \cdot Str \cdot d_{c}^{2} }}} \right)^{2} }}\prod {K(t)}^{2} \\ + \sum\limits_{l = - \infty }^{\infty } {w^{\prime}(n_{0} ) - \eta^{\prime}_{sjh} \frac{\partial J}{{\partial w^{\prime}_{sjh} }}\left( {\pi T_{B} \left( {f - \frac{l}{{NT_{C} }}} \right)} \right)} \\ \end{gathered}$$

(7)

where,

$$K(t)^{{}} = T_{C}^{2} N\sin c^{2} \left( {\pi fT_{C} } \right) + R_{dc} (\frac{{I^{\prime}}}{\omega \cdot I})$$

(8)

$$T_{c} = \frac{1}{\sqrt N }\sum\limits_{n = 0}^{N - 1} {x_{n} \exp \left( { - j2\pi fnT_{c} } \right)}$$

(9)

In the above formula, $T_{C}$ is the sampling threshold of the direct predictor of the new power system net load, and $f$ is the dynamic discrete sequence reconstruction model of risk management and control of load uncertainty. It is known that the random probability density function is obtained as $R = R_{dc} + R(l)$ by analyzing the uncertain scenario for short-term probability prediction parameters:

$$\begin{gathered} P(U\left| \beta \right.) = (2\pi )^{ - d/2} \left| {\sum {\left[ {e^{b}_{m - 1} \left( {n - 1} \right)} \right]} } \right|^{ - 1/2} \\ \times \exp \left[ { - \frac{1}{2}\left[ {e^{f}_{m - 1} \left( n \right)} \right]^{T} \sum\nolimits_{k}^{ - 1} {(U - u_{k} )} } \right] \\ \end{gathered}$$

(10)

wherein, $e^{f}_{m - 1} \left( n \right)$ is the statistical probability density characteristic quantity of energy network load energy storage on the load side of the new power system. The dynamic parameters of the energy network load storage on the load side of the new power system are collected using local bus transmission technology. The collected dynamic parameters are fused with the energy distribution state parameters to extract the state characteristic quantities of the energy network load storage [15].

Multi-objective cooperative scheduling optimization of energy network load and storage

Simulated annealing algorithm design

A simulated annealing algorithm is used to achieve load feature fusion and adaptive scheduling of load-side energy networks of power systems and to extract the spectral eigenquantities of the dynamic parameters of loads [16]. The necessity of using a simulated annealing algorithm is that it enables load feature fusion and adaptive scheduling of the load-side energy networks of power systems. The spectral eigenvolumes of the load dynamic parameters are extracted to achieve multi-objective optimization of the dynamic loads. Its innovations include being used for the first time to solve the multi-objective co-scheduling problem of load storage in a new power system load-side energy network, combining the local bus transmission technology and energy allocation state parameters, and the real-time acquisition and fusion of dynamic parameters. It also provides a basis for subsequent dynamic scheduling through load-characteristic fusion and adaptive scheduling processing.

By using the deterministic point prediction method, the concept set of energy storage parameter fusion of the energy network on the load side of the power system is obtained as follows:

$$E(e^{ - sX} ) = \exp ( - \frac{{\sigma^{\alpha } s^{\alpha } }}{\cos (\pi \alpha /2)})$$

(11)

Based on the analysis of energy-saving conditions and the results of short-term probabilistic forecasting of user loads, an analytical model with high load stochasticity and volatility is established. The conditional probability density function for joint load dynamic monitoring is obtained:

$$E_{n} = \sum\limits_{m = - \infty }^{\infty } {[K(r)G(n - m)]^{2} } + e^{f}_{m - 1} (n)e^{b}_{m - 1} (n - 1)$$

(12)

The fusion of neighboring submodules and design of the simulated annealing algorithm are utilized. Obtain the spatial spectral density function predicted by dynamic optimization iterative deep learning during simulated annealing:

$$M(h) = \left\{ \begin{gathered} \sigma \sum\limits_{i = 1}^{p} {a_{i} } R_{x} (m - i),m = 1,2,3.....p \hfill \\ \sigma \prod {Mh} \times \frac{{E_{j} }}{{\sum\limits_{j = 1}^{J + 1} {E_{j} } }},j = J + 1 \hfill \\ \end{gathered} \right.$$

(13)

where, $N$ represents the iterative length of simulated annealing on the load side of the new power system, and $J$ is the linear high-order statistical frequency of quantile regression analysis. The $p + l$ parameters of the random time series are used for the linear combination processing of the new power system load, and the spectral distribution results of photovoltaic difference model and load difference are obtained as follows:

$$P(m) = \sum\limits_{i = 1}^{p} {a_{i} } R_{x} (i) + G^{2} ,m = 0,\quad m{ = 1,2,}...{,}M + 1,$$

(14)

where $G = e^{ - \delta } + \sum\limits_{i = 1}^{p} {a_{i} } R_{x} (i)$, $m > 0$. According to the dynamic update of the simulated annealing of the load-side cross iteration of the new power system, the statistical characteristic sequence of the multi-objective cooperative scheduling of load forecasting is obtained as follows:

$$K(t) = {\raise0.7ex\hbox{$R$} \!\mathord{\left/ {\vphantom {R {\rho_{m - 1} }}}\right.\kern-0pt} \!\lower0.7ex\hbox{${\rho_{m - 1} }$}} = \frac{{P_{L} }}{{P_{E} }} + \sum\limits_{i = 1}^{c} {\mu_{ik} = 1,} k = 1,2, \cdots ,n$$

(15)

where, $k$ is the gray sequence of energy network load energy storage on the load side of new power system. A simulated annealing algorithm is used to realize load feature fusion and adaptive scheduling of load-side energy networks in power systems to extract the spectral features of the load dynamic parameters. The dynamic scheduling method using simulated annealing is used to achieve multi-objective optimization of dynamic loads in energy networks [17].

Multi-objective collaborative scheduling optimization of energy network load and energy storage

The variance of the forecast residuals at different time points is calculated using the distributed optimal control technique and simulated annealing technique for energy microgrid clusters. Based on the analysis of multivariate constraint indicators, the dynamic scheduling of the energy network load storage on the load side of the new power system is realized. The Gaussian probability density distribution function of the energy scheduling distribution is represented by the joint spectral feature:

$$\begin{gathered} Q(t) = \left( {\begin{array}{*{20}c} {M_{x} (0)} & {M_{x} (1)} & {......} & {M_{x} (p - 1)} \\ {M_{x} (1)} & {M_{x} (0)} & {......} & {M_{x} (p - 2)} \\ {.....} & {......} & {......} & {.......} \\ {M_{x} (p - 1)} & {M_{x} (p - 2)} & {......} & {M_{x} (0)} \\ \end{array} } \right) \\ + \left( {\begin{array}{*{20}c} 1 & {q_{1} } & \cdots & {q_{1}^{m - 1} } \\ 1 & {q_{2} } & \cdots & {q_{2}^{m - 1} } \\ \vdots & \vdots & \ddots & \vdots \\ 1 & {q_{m} } & \cdots & {q_{m}^{m - 1} } \\ \end{array} } \right) \\ \end{gathered}$$

(16)

The point prediction model is utilized to obtain the point prediction results to construct a new power system load-side energy scheduling clustering model. The variance of the point prediction residuals for different periods is calculated, and the set of energy consumption feature distributions is obtained as follows:

$$\begin{gathered} E = R_{x} (0) + 2\sum\limits_{i = 1}^{p} {a_{i} } R_{x} (i) + \sum\limits_{i = 1}^{p} {a_{i} \left[ { - R_{x} (i)} \right]} \\ = R_{x} (0) + \sum\limits_{i = 1}^{p} {a_{i} } R_{x} (i) \\ \end{gathered}$$

(17)

Assuming that the errors are distributed, a simulated annealing algorithm is used for the design. The joint analysis matrix output from the point prediction co-scheduling model and residual prediction co-scheduling model satisfy the following requirements:

$$M = R_{x} (0) + \sum\limits_{i = 1}^{p} {a_{i} R_{x} } (i) = G^{2} \left| {\begin{array}{*{20}c} {t_{{i_{1} ,j_{1} }} } & {t_{{i_{1} ,j_{2} }} } & \cdots & {t_{{i_{1} ,j_{k} }} } \\ {t_{{i_{2} ,j_{1} }} } & {t_{{i_{2} ,j_{2} }} } & \cdots & {t_{{i_{2} ,j_{k} }} } \\ \vdots & \vdots & \ddots & \vdots \\ {t_{{i_{k} ,j_{1} }} } & {t_{{i_{k} ,j_{2} }} } & \cdots & {t_{{i_{k} ,j_{{_{k} }} }} } \\ \end{array} } \right|$$

(18)

A continuous energy-saving parametric analysis model for load storage in energy networks on the load side of a new power system is constructed. Dynamic load multi-objective optimization of the energy network was realized using the dynamic scheduling method of simulated annealing. According to the multi-objective optimization and co-optimization results of simulated annealing, the optimization application of the simulated annealing algorithm in multi-objective co-optimization scheduling of the load storage of the energy network on the load side of the new power system is realized. The realization process is illustrated in Fig. 2.

Fig. 2

Implementation process of energy network load energy storage calculation on load side of new power system

Energy storage influences the model by providing a way to balance the dynamic loads. It allows for the absorption and release of energy, which in turn affects the overall stability and efficiency of the energy network. In the equation, storage affects parameters, such as the energy network load storage parameters and balancing parameters. When it interacts with the load, it changes the distribution of frequency and voltage parameters.

Experimental test analysis

To verify the method’s performance in accurately calculating new power system load side energy network load storage, experimental tests are conducted. Matlab was used to design the load storage calculation algorithm. In the experiment, VMD and quantile generated the load prediction quantile directly. Energy parameters were collected at four time points (0:00, 6:00, 12:00, and 18:00). Considering energy savings, the model optimizes load and storage scheduling, reducing energy waste and improving system efficiency. The Newton–Raphson method was used in the analytical control model to adjust and optimize parameters using the principle of energy conservation, ensuring efficient load and storage management. The specific parameter settings are shown in Table 1.

Table 1 Specific situation of parameter settings

Based on these parameter settings, a simulated annealing algorithm is programmed to implement a multi-objective co-scheduling optimization process for energy network loads and energy storage in a MATLAB environment. The simulated annealing algorithm performs iterative calculations based on the set parameters, and in each iteration, decides whether or not to accept a new solution based on the evaluation of the current solution and the neighboring solutions, as well as the temperature parameters, and seeks to optimize step by step in order to achieve the multi-objective optimization of load feature fusion, adaptive scheduling processing, and dynamic loads, and ultimately obtains the optimization results.

Given the current limit value of 0.32pu, the acceleration process takes about 0.17 s, and the motor working conditions of the power system are all 0.6pu speed and 0.9pu torque. According to the above parameter settings, the load time series is first given, as shown in Fig. 3.

Fig. 3

Load time series

A simulated annealing algorithm is used to realize load feature fusion and adaptive scheduling of load-side energy networks in power systems, and the spectral feature quantities of the load dynamic parameters are extracted. Dynamic load multi-objective optimization of the energy network is achieved using the simulated annealing dynamic scheduling method. The spectral feature distribution results of dynamic load parameters are shown in Fig. 4.

Fig. 4

Spectral characteristic distribution results of load dynamic parameters

The time series of load energy storage distribution of power system depicted in Fig. 4 exhibits disturbances, resulting in diminished anti-interference capabilities and convergence. The method proposed in this paper is used for co-scheduling, and a comparison of the dynamic load scheduling results of the power system is shown in Fig. 5. The advantage of the proposed method lies in its ability to effectively address the issue of disturbed load storage distribution in the power system, thereby enhancing anti-interference capabilities and convergence. By using this method for collaborative scheduling, a comparative analysis was conducted with the traditional method based on fixed priority sequential scheduling as a control group, yielding superior dynamic load scheduling results. This outcome can be attributed to the advanced algorithms and techniques incorporated in the proposed method, which enable accurate and effective management and optimization of the load and storage of the source network on the load side of the power system.

Fig. 5

Comparison of multi-objective cooperative scheduling of load and storage of source network on load side of power system

From the analysis of Fig. 5, it is known that the method in this study has good convergence and high stability for multi-objective cooperative scheduling of load and storage of the source network on the load side of the power system. Test the accuracy of different methods for energy network load energy storage scheduling on the load side of a new power system, and get the comparison results as shown in Table 2. By analyzing Table 2, it is known that the method proposed in this study is more accurate for the energy network load energy of a new power system. The superiority of the methodology proposed in this study stems from its ability to achieve high convergence accuracy and stability in multi-objective co-scheduling of load and storage on the load side of new power systems. This is attributed to the effective use of simulated annealing for load feature fusion and adaptive scheduling as well as the accurate computation of energy network load storage via VMD and quantile-based load forecasting. Together, these factors contribute to the excellent performance of the method, with significantly higher control convergence accuracy compared with other reference methods.

Table 2 Comparison results of convergence accuracy of different methods

To validate the proposed method’s effectiveness and superiority, the load scheduling stability index is used to evaluate performance. This index measures the deviation between load scheduling results and actual demand, as well as the fluctuation of this deviation. A lower LSSI indicates better system stability. The power system is divided into multiple time periods, with load scheduling performed once per period. The load scheduling data for each period is recorded, and the comparison results are shown in Table 3 Table 3 analysis shows that this method has clear advantages over other methods. The scheduling efficiency improvement is 0.25%, higher than other methods, indicating enhanced system operation efficiency. The load scheduling stability index is 0.025, significantly lower than other methods, implying good system stability. The scheduling time is only 200 s, lower than other methods, showing a high scheduling efficiency. The energy waste ratio is 2.0%, which is lower than other reference methods, proving effective energy waste reduction, thus fully verifying the proposed method’s effectiveness and superiority in multi-objective coordinated scheduling of load and energy storage.

Table 3 Comparison results of convergence accuracy of different methods

Conclusions

This paper proposes a novel method for multi-objective cooperative scheduling of energy network load storage based on the simulated annealing algorithm. Local bus transmission technology is used to collect dynamic parameters of energy network load storage on the load side of the new power system. The collected load dynamic parameters are integrated with energy distribution state parameters to extract the state characteristics of energy network load storage. The simulated annealing algorithm is used to achieve load characteristic fusion and adaptive scheduling processing of the load side energy network of the power system, and the spectral characteristics of the load dynamic parameters are extracted. The dynamic scheduling method of simulated annealing is used to accomplish the multi-objective optimization of the dynamic load of the energy network. Based on the collaborative optimization results of simulated annealing, the optimized application of the simulated annealing algorithm in the multi-objective collaborative scheduling of load and storage of new power system is achieved. The experimental results demonstrate that the method exhibits effective multi-objective integration and robust optimization ability in the scheduling process, enhancing the load balancing scheduling and adaptive control ability of the power system. The complexity of the net load is reduced, and the characteristics are more distinct and readily interpretable and learnable. Consequently, the prediction accuracy is improved to certain extent. Finally, the probability density prediction results and interval prediction results of the proposed prediction method are presented, and the prediction errors are analyzed.