Node synchronous playing method and system for distributed player

Abstract

The invention provides a node synchronous playing method and system for a distributed player, and relates to the field of video playing, and the method comprises the steps: carrying out the structural coding and node state clustering analysis of the state information of each node in a node cluster, obtaining the maximum one in a set of node performance deviation values through the calculation of the node performance deviation values, and obtaining the maximum one in the set of node performance deviation values; the corresponding node is used as a master node, and the rest are slave nodes. Then, acquiring basic information of the video through the main node, calculating playing frame information and sending a playing state; then, the target frame of the slave node is adjusted through the network delay between the slave node and the master node, and frame-level synchronization is maintained; thus, even in the face of local network jitter or transient communication interruption, it can be guaranteed that the nodes operate based on unified logic, visual coherence and expressive force in scenes such as light show and the like are improved, and accurate picture synchronization between large-scale node clusters is achieved.

Description

Technical Field

[0001] The present application relates to the field of video playback, and more particularly, in an embodiment of the present application, to a node synchronization playback method and system for a distributed player. Background Art

[0002] The rapid development of IoT and lighting technologies has driven the widespread global adoption of landscape lighting and light shows. This not only adds a unique charm to urban nightscapes but also provides a unique visual experience for various commercial and cultural events.

[0003] However, as market demand continues to rise, its application scenarios are becoming increasingly complex. Especially in large-scale lighting system applications, ensuring the precise synchronization of program images across multiple nodes has become a critical issue that needs to be addressed. In actual operation, network latency, clock asynchrony caused by long-term device operation, and node communication disconnections caused by various unexpected situations frequently occur. These issues directly lead to asynchronous video program playback between nodes, severely damaging the overall presentation of the light show and significantly reducing the expressiveness of the images. This significantly impacts the audience's viewing experience and prevents them from fully appreciating the unique charm of the light show.

[0004] Therefore, a node synchronization playback solution for distributed players is desired. Summary of the Invention

[0005] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides a node synchronization playback method and system for a distributed player, which performs structured coding and node status clustering analysis on the status information of each node in the node cluster, and then performs a node performance offset value to obtain the largest one in the set of node performance offset values, and uses the corresponding node as the master node, and the rest as slave nodes. Then, the master node obtains the basic information of the video, calculates the playback frame information and sends the playback status. Then, the target frame itself is adjusted by the network delay between the slave node and the master node to maintain frame-level synchronization. In this way, even in the face of local network jitter or short-term communication interruption, it can ensure that the nodes operate based on a unified logic, improve the visual coherence and expressiveness in scenes such as light shows, and achieve accurate picture synchronization between large-scale node clusters.

[0006] According to one aspect of the present application, a node synchronization playback method for a distributed player is provided, comprising:

[0007] Designate a master node from a node cluster and designate other nodes in the node cluster as slave nodes;

[0008] The master node obtains basic information of the video program and calculates the playback frame information based on the current playback progress;

[0009] The master node sends playback status information to the slave nodes via broadcast;

[0010] After receiving the playback status information, the slave node calculates the network delay between the slave node and the master node based on the playback status information and the round-trip time of the data packet;

[0011] The slave node calculates the target frame that it should currently play based on the playback frame information and the network delay.

[0012] According to another aspect of the present application, a node synchronization playback system for a distributed player is provided, comprising:

[0013] A node cluster designation module is used to designate a master node from a node cluster and designate other nodes in the node cluster as slave nodes;

[0014] The master node playback frame information calculation module is used to obtain basic information of the video program through the master node and calculate the playback frame information based on the current playback progress;

[0015] The playback status information sending module is used to send playback status information to the slave nodes through the master node in a broadcast manner;

[0016] A network delay calculation module is used to calculate the network delay between the slave node and the master node based on the playback status information and the round-trip time of the data packet after receiving the playback status information;

[0017] The current playback target frame calculation module is used to calculate the target frame that its own slave node should currently play based on the playback frame information and the network delay.

[0018] Compared with the prior art, the present application provides a node synchronization playback method and system for a distributed player, which performs structured coding and node status clustering analysis on the status information of each node in the node cluster, and then performs node performance offset value to obtain the largest one in the set of node performance offset values, and uses the corresponding node as the master node, and the rest as slave nodes. Then, the master node obtains the basic information of the video, calculates the playback frame information and sends the playback status. Then, the target frame itself is adjusted by the network delay between the slave node and the master node to maintain frame-level synchronization. In this way, even in the face of local network jitter or short-term communication interruption, it can ensure that the nodes operate based on a unified logic, improve the visual coherence and expressiveness in scenes such as light shows, and achieve accurate picture synchronization between large-scale node clusters. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The above and other purposes, features, and advantages of the present application will become more apparent through a more detailed description of the embodiments of the present application in conjunction with the accompanying drawings. The accompanying drawings are intended to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the drawings, the same reference numerals generally represent the same components or steps.

[0020] Figure 1 This is a flowchart of a node synchronization playback method for a distributed player according to an embodiment of the present application.

[0021] Figure 2 This is a flowchart of automatically electing a node from the node cluster as the master node based on the node status information of each node in the node cluster in the node synchronization playback method for a distributed player according to an embodiment of the present application.

[0022] Figure 3 This is a data flow diagram for automatically electing a node from the node cluster as the master node based on the node status information of each node in the node cluster in the node synchronization playback method for a distributed player according to an embodiment of the present application.

[0023] Figure 4 A flowchart of a node state clustering analysis for autonomously transforming a set of structured coding vectors of node state information to obtain a node state cluster modeling coding vector in a node synchronization playback method for a distributed player according to an embodiment of the present application.

[0024] Figure 5This is a flowchart of calculating the clustering correction incremental operator of each node state information structured coding vector in the set of node state information structured coding vectors based on the node state information structured initial linear clustering center coding vector in the node synchronization playback method for a distributed player according to an embodiment of the present application to obtain a set of node state information structured clustering correction incremental operators.

[0025] Figure 6 This is a system block diagram of a node synchronous playback system for a distributed player according to an embodiment of the present application. DETAILED DESCRIPTION

[0026] Various exemplary embodiments, features, and aspects of the present application will be described in detail below with reference to the accompanying drawings. The same reference numerals in the accompanying drawings represent elements with the same or similar functions. Although various aspects of the embodiments are shown in the accompanying drawings, the drawings are not necessarily drawn to scale unless otherwise indicated.

[0027] The word “exemplary” is used exclusively herein to mean “serving as an example, example, or illustration.” Any embodiment described herein as “exemplary” is not necessarily to be construed as preferred or advantageous over other embodiments.

[0028] In addition, numerous specific details are provided in the following detailed description to better illustrate the present application. Those skilled in the art will appreciate that the present application can be practiced without certain specific details. In some instances, methods, means, components, and circuits well known to those skilled in the art are not described in detail in order to highlight the main purpose of the present application.

[0029] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of the technical features being referred to. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features. Throughout the description of this application, "plurality" means two or more, unless otherwise specifically defined.

[0030] The rapid development of the Internet of Things and lighting technology has promoted the widespread application of landscape lighting and light shows around the world, adding charm to urban nightscapes and providing a unique visual experience for commercial and cultural activities. However, with the growth of market demand and the complexity of application scenarios, ensuring the precise synchronization of multiple nodes in large-scale lighting systems has become a key challenge. Problems such as network latency, device clock asynchrony, and node communication disconnection are frequent, resulting in asynchronous video playback, which damages the overall effect and expressiveness of the light show, seriously affecting the audience experience, and hindering the display of the unique charm of the light show. In response to the above technical problems, this application proposes a node synchronization playback method for a distributed player. Figure 1FIG is a flow chart of a node synchronization playback method for a distributed player according to an embodiment of the present application. Figure 1 As shown, the node synchronization playback method for a distributed player according to an embodiment of the present application includes: S110, specifying a master node from a node cluster, and specifying other nodes in the node cluster as slave nodes; S120, the master node obtains basic information of the video program and calculates playback frame information based on the current playback progress; S130, the master node sends playback status information to the slave node via broadcast; S140, after receiving the playback status information, the slave node calculates the network delay between the slave node and the master node based on the playback status information and the round-trip time of the data packet; S150, the slave node calculates the target frame that it should currently play based on the playback frame information and the network delay.

[0031] Specifically, this solution builds a cluster architecture by dynamically electing a master node. This allows for rapid, automatic failover in the event of a master node failure, avoiding the long synchronization interruptions associated with master node failures in traditional solutions and enhancing fault tolerance for unexpected node disconnections. The master node continuously broadcasts playback status information, including key data such as playback frames and timestamps, to slave nodes. This allows slave nodes to not only obtain the master node's playback progress benchmark in real time but also accurately calculate network transmission latency based on the round-trip time of data packets, overcoming the latency estimation errors often associated with one-way timing synchronization due to network fluctuations. Furthermore, slave nodes convert network latency and a preset custom delay time into a frame offset value. By dynamically adjusting the playback timing of target frames, they mitigate frame skew caused by accumulated hardware clock errors between devices and provide differentiated synchronization strategies for nodes in diverse network environments. This multi-layered coordination mechanism ensures that even in large-scale node clusters, even with localized network jitter or brief communication interruptions, each node maintains frame-level synchronization based on a unified playback logic, significantly enhancing the coherence and expressiveness of visual effects in scenes such as light shows.

[0032] In the above-mentioned node synchronization playback method for a distributed player, step S110, which designates a master node from a node cluster and designates other nodes in the node cluster as slave nodes, includes automatically electing a node from the node cluster as the master node based on node status information of each node in the node cluster. It should be understood that in the actual operation of large-scale lighting systems, network conditions and device status may change at any time. Electing a master node based on node status information can comprehensively consider the current status of each node, such as its processing capacity, network connection stability, and the presence of faults. This ensures that the elected master node is better equipped to coordinate and control the synchronized playback of the entire cluster under the current circumstances, adapting to complex and changing application scenarios. Traditional solutions typically use manual or fixed master node designation. This static mechanism has a fatal flaw: if the pre-set master node fails due to hardware failure, network fluctuations, or sudden excessive load, it loses the ability to generate and distribute global synchronization instructions, causing the slave nodes to enter a disordered state due to a lack of a reference reference. Especially in landscape lighting scenarios deployed across regions, nodes in different geographical locations may experience regional network fluctuations due to differences in metropolitan area network transmission paths. If the master node cannot dynamically migrate to the optimal location node, the edge node will have significant synchronization lag due to the long instruction transmission path, and even cause a cascading collapse of the entire system due to network paralysis in the area where the master node is located.

[0033] Based on this, in automatically electing a node from the node cluster as the master node based on the node status information of each node in the node cluster, the technical concept of the present application is to first perform structured encoding on the real-time status information of each node (including processing capacity, network performance, load conditions, etc.), and convert it into a set of vectors that can be quantified and analyzed; then, through cluster analysis of autonomous evolution, an optimal clustering feature model of the node status is established from a global perspective, and a coding vector representing the overall optimal operating state of the cluster is generated; then, the degree of performance deviation of each node state vector from the ideal model is calculated, and finally the node with the largest deviation value is selected as the master node. This technical concept effectively solves the single point failure risk and regional network fluctuation sensitivity problems under the traditional fixed master node architecture: when the original master node experiences performance degradation or the regional network is abnormal, it can quickly identify and migrate the master node responsibilities to nodes with better performance and shorter network paths, avoiding a surge in command transmission delays or loss of global synchronization benchmarks due to master node failure; at the same time, through intelligent clustering evaluation of node status, it can adaptively adjust the master node selection strategy, and give priority to nodes with high network topology centrality and strong communication stability in cross-regional deployment scenarios, significantly reducing the risk of synchronization lag at the edge nodes, thereby ensuring frame-level precise synchronization of the playback timing of each node in the large-scale light show system and system-level disaster recovery capabilities.

[0034] Figure 2 This is a flowchart of automatically electing a node from the node cluster as the master node based on the node status information of each node in the node cluster in the node synchronization playback method for a distributed player according to an embodiment of the present application. Figure 3 This is a data flow diagram of automatically electing a node from the node cluster as the master node based on the node status information of each node in the node cluster in the node synchronous playback method for distributed players according to an embodiment of the present application. Figure 2 and Figure 3 As shown, in an embodiment of the present application, the step S110, based on the node status information of each node in the node cluster, automatically elects a node from the node cluster as the master node, including: S111, structured encoding of the node status information of each node to obtain a set of node status information structured encoding vectors; S112, autonomously transformed node status clustering analysis of the set of node status information structured encoding vectors to obtain a node status cluster modeling encoding vector; S113, based on the node performance offset value of each node status information structured encoding vector in the set of node status information structured encoding vectors relative to the node status cluster modeling encoding vector to obtain a set of node performance offset values; S114, taking the node corresponding to the largest node performance offset value in the set as the master node.

[0035] Specifically, the step S111 is to perform structured encoding on the node status information of each node to obtain a set of structured encoding vectors of node status information. It should be understood that in the technical solution of the present application, first of all, considering that the node status information usually contains multi-dimensional heterogeneous data (such as CPU load, memory occupancy, network delay, communication success rate, etc.), these indicators have different dimensions, numerical ranges and change frequencies, and the traditional master node election algorithm is difficult to directly perform effective fusion and quantitative evaluation on the original heterogeneous data. For example, network delay is measured in milliseconds, while CPU load is a percentage value. There are significant differences in the numerical magnitude and fluctuation characteristics between the two. If a simple weighted summation method is used to compare node status, the key performance characteristics may be weakened or obscured due to improper indicator normalization methods. In particular, in cross-regional deployment scenarios, nodes in different geographical locations may present completely different state feature combinations due to network topology differences (such as central nodes have low latency but high load, edge nodes have low load but large latency fluctuations). The unstructured characteristics of this multi-dimensional data make it impossible for traditional election mechanisms to accurately capture the comprehensive performance differences between nodes, which easily causes the master node selection to deviate from the actual optimal solution. Based on this, the present application performs structured encoding on the node status information of each node to obtain a set of structured encoding vectors of the node status information. That is, by establishing a unified data representation framework, the multi-dimensional operating parameters of each node are converted into a comparable vector space representation. Specifically, this step uses feature engineering methods to standardize and align the original state data, eliminate the influence of different dimensions on the analysis results, and extract the nonlinear correlation between key performance indicators through feature embedding technology, and finally generate a structured encoding vector reflecting the comprehensive state of the node. For example, for the time series volatility indicator of network delay, the encoding process not only considers the current instantaneous value, but also counts its historical stability characteristics through a sliding window; for discrete state parameters (such as fault marks), a unique hot encoding method is used to map them to a continuous vector space. This structured processing enables the originally dispersed and heterogeneous node status data to be converted into high-dimensional vectors with a unified mathematical expression, laying a data foundation for subsequent intelligent analysis.

[0036] Figure 4 The flowchart of the node state cluster modeling coding vector is obtained by performing autonomous transformation of the node state cluster analysis on the set of the node state information structured coding vectors in the node synchronous playback method for distributed players according to the embodiment of the present application. Figure 4As shown, in an embodiment of the present application, the step S112, which performs autonomously evolved node state clustering analysis on the set of the node state information structured coding vectors to obtain a node state clustering modeling coding vector, includes: S1121, performing linear clustering analysis on the set of the node state information structured coding vectors to obtain a node state information structured initial linear clustering center coding vector; S1122, based on the node state information structured initial linear clustering center coding vector, calculating the clustering correction incremental operator of each node state information structured coding vector in the set of the node state information structured coding vector to obtain a set of node state information structured clustering correction incremental operators; S1123, based on the set of node state information structured clustering correction incremental operators, calculating the node state information structured linear clustering correction component coding vector of the set of the node state information structured coding vector; S1124, fusing the node state information structured linear clustering correction component coding vector and the node state information structured initial linear clustering center coding vector to obtain the node state clustering modeling coding vector. It should be understandable that, considering that the structured coding vector of the node state information solves the problem of data heterogeneity, it is difficult to capture the complex correlations between nodes by relying solely on linear clustering methods. Traditional linear clustering can only identify explicit linear distribution patterns, while the actual node state often presents nonlinear characteristics (such as the dynamic coupling of network delay and CPU load, and implicit topological associations between regional nodes). For example, in a cross-regional deployment scenario, the central node may have a nonlinear negative correlation with the network delay and processing power due to the load balancing strategy, while the low load state of the edge node may mask the delay jitter risk introduced by long-distance transmission. If clustering is performed based only on the linear distance metric, nodes with different performance characteristics may be mistakenly classified into the same category, resulting in ignoring key performance shortcomings when electing the master node, thereby affecting the synchronization accuracy. Therefore, the present application performs autonomously evolved node state clustering analysis on the set of structured coding vectors of the node state information to obtain a node state clustering modeling coding vector.

[0037] Specifically, linear clustering is first used to extract the principal component features of the node state, constructing an initial cluster center encoding vector that represents the global linear structure of the cluster, providing a basic framework for subsequent analysis. Next, deep collaborative implicit coding is used to explore the nonlinear interactions between the node state vector and the initial cluster center. For example, this captures the dynamic balance between low load and high latency at edge nodes, generating a deep implicit vector that reflects the local characteristics of the node. Subsequently, a cluster correction increment operator is used to perform nonlinear corrections on the linear clustering results to reflect the actual communication quality. Finally, the linear and nonlinear features are integrated to generate a node state cluster modeling encoding vector. This vector not only retains the macroscopic distribution characteristics of the node state but also embeds dynamic correlation information at the micro level. It provides a global reference standard for evaluating the degree of deviation of each node from the overall optimal state.

[0038] Specifically, in step S1121, a linear cluster analysis is performed on the set of the node state information structured coding vectors to obtain the node state information structured initial linear cluster center coding vector, which is expressed as follows using the linear cluster analysis formula:

[0039] X={x1,x2,...,x i ,...,x n}

[0040]

[0041] Among them, X is the set of structured coding vectors of node state information, x1, x2, x i and x n are the first, second, i-th and n-th node state information structured coding vectors in the set of node state information structured coding vectors, n is the number of vectors in X, x cis the structured initial linear cluster center encoding vector of node state information. It should be understood that node state information is essentially a high-dimensional space composed of multidimensional heterogeneous data. This data includes parameters with varying dimensions, fluctuation frequencies, and semantic characteristics, such as network latency, CPU load, and memory usage. Traditional methods struggle to effectively integrate and quantitatively evaluate this heterogeneous data. However, linear cluster analysis, through unsupervised learning, identifies natural grouping patterns in the data within this high-dimensional space. By iteratively optimizing to minimize the sum of squared intra-cluster distances, it maps the dispersed node state information into a linear subspace defined by the cluster centers. This process essentially performs a low-rank approximation on the original feature space, extracting the most significant linear principal components from the data to form a global feature skeleton. This skeleton serves as the foundational anchor for subsequent nonlinear feature gain aggregation. It not only provides a structured reference framework for deep learning models but also explicitly preserves the main linear patterns in the dataset by decoupling linear and nonlinear components. This allows subsequent steps to more efficiently focus on mining complex nonlinear relationships between node states (such as the dynamic coupling between network latency and processing power), thereby providing a feature representation that combines a global perspective with local details for dynamic master node election. Finally, the obtained structured initial linear cluster center encoding vector of node status information realizes the quantitative evaluation of node status differences by constructing a benchmark paradigm for feature representation, ensuring that the master node election mechanism can adapt to dynamic changes in network topology and node load fluctuations, laying the foundation for precise synchronization of large-scale clusters.

[0042] Figure 5 This is a flowchart of calculating the cluster correction increment operator of each node state information structured coding vector in the set of node state information structured coding vectors based on the node state information structured initial linear cluster center coding vector in the node state information synchronized playback method for distributed players according to an embodiment of the present application to obtain a set of node state information structured cluster correction increment operators. Figure 5As shown, in an embodiment of the present application, the step S1122, based on the node state information structured initial linear clustering center coding vector, calculates the clustering correction incremental operator of each node state information structured coding vector in the set of node state information structured coding vectors to obtain a set of node state information structured clustering correction incremental operators, including: S1122-1, constructing a node state information structured deep collaborative implicit coding vector between each node state information structured coding vector in the set of node state information structured coding vectors and the node state information structured initial linear clustering center coding vector; S1122-2, based on the node state information structured deep collaborative implicit coding vector, calculates the characteristic clustering correction incremental operator of each node state information structured coding vector in the set of node state information structured coding vectors relative to the node state information structured initial linear clustering center coding vector to obtain a set of node state information structured clustering correction incremental operators.

[0043] Specifically, step S1122-1 constructs a node state information structured deep collaborative implicit coding vector between each node state information structured coding vector in the set of node state information structured coding vectors and the node state information structured initial linear clustering center coding vector, and the node state information structured deep collaborative implicit coding formula is expressed as:

[0044] r i =Sigmoid{W i [concat(x i ;x c )]+b i}

[0045] Among them, concat(·;·) is the connection operation, W i is the i-th learnable collaborative weight matrix among multiple learnable collaborative weight matrices, b i is the i-th collaborative bias vector among multiple collaborative bias vectors, Sigmoid is the Sigmoid activation function, r i is x i and x cThe node state information structured deep collaborative implicit encoding vector between the node state information structured encoding vector and the node state information structured initial linear cluster center encoding vector is constructed through a deep learning model. It should be understood that while linear clustering analysis can extract preliminary linear principal components of node state data, it is essentially a dimensionality reduction projection of high-dimensional features and cannot capture the complex dynamic coupling relationships and nonlinear correlation patterns between node states. For example, network latency and device load may exhibit a nonlinear negative correlation, while the low load status of edge nodes may mask the latency jitter risk introduced by long-distance transmission. Relying solely on linear clustering results may cause key performance shortcomings to be overlooked during master node election, thereby affecting synchronization accuracy. To this end, a deep learning model is used to construct a node state information structured deep collaborative implicit encoding vector between the node state information structured encoding vector and the node state information structured initial linear cluster center encoding vector. Essentially, this model uses nonlinear transformations to mine the contextual information of the node state information structured encoding vector relative to the global linear cluster structure. This process utilizes a deep collaborative mechanism, so that the generation of the node state information structured deep collaborative implicit encoding vector is driven by both the node state characteristics and the global linear cluster structure, thereby establishing a deep correlation between node states and the optimal cluster distribution pattern in the feature space. By combining the Sigmoid activation function with incremental computation at the granularity of eigenvalues, the model can learn the nonlinear distribution of node states in high-dimensional space and extract implicit features reflecting complex relationships between nodes, such as the temporal correlation between node load fluctuations and communication stability, and the implicit impact of regional network topology on latency. This deep collaborative encoding not only overcomes the shortcomings of linear clustering in representing nonlinear features, but also forms a semantically rich high-order feature representation by fusing the global positioning information and local dynamic characteristics of node states. This provides refined input for the subsequent computation of the cluster correction incremental operator, enabling the master node election mechanism to more accurately identify nodes with the best overall performance, thereby improving the synchronization robustness and disaster recovery capabilities of large-scale clusters in dynamic network environments.

[0046] In an embodiment of the present application, the step S1122-2, based on the node state information structured deep collaborative implicit coding vector, calculates the characteristic clustering correction increment operator of each node state information structured coding vector in the set of the node state information structured coding vector relative to the node state information structured initial linear clustering center coding vector to obtain the set of the node state information structured clustering correction increment operator, including: S1122-21, performing nonlinear activation processing based on the Sigmoid function on the node state information structured initial linear clustering center coding vector to obtain the node state information structured initial linear clustering center activation coding vector; S1122-22 , perform incremental calculation based on eigenvalue granularity on the node state information structured initial linear clustering center activation coding vector and the node state information structured deep collaborative implicit coding vector corresponding to the node state information structured coding vector to obtain the node state information structured clustering increment; S1122-23, perform dynamic relaxation parameter collaborative resonance coupling optimization on the node state information structured clustering increment to obtain the node state information structured clustering optimization increment; S1122-24, perform softmax-based normalization processing on the node state information structured clustering optimization increment to obtain the node state information structured clustering correction increment operator corresponding to the node state information structured coding vector.

[0047] Specifically, the step S1122-2 is expressed as follows:

[0048]

[0049] Among them, v c is the node state information structured initial linear cluster center activation encoding vector, is r i The kth eigenvalue in , log2 is the logarithmic function value with base 2, It is v c The kth eigenvalue in, D is r i and v c The length of the vector, and r i and v c The same length, λ i is x i The corresponding node state information structured clustering increment, <,> represents the vector inner product, Σ i is x i The corresponding node state information structured eigenvalue, σ(x i -r i ) 2 Represents a vector (x i -r i), L is the variance of the vector (x i -r i )’s length,∈ i is x i The corresponding node state information structured intrinsic scattering coefficient, Γ i is x i The corresponding node state information structured clustering dynamic relaxation parameter, λ' i is x i The corresponding node state information structured clustering optimization increment, softmax is the normalization function, ε i is x iCorresponding node state information structured clustering correction increment operator. It should be understood that in the node synchronization playback method of the distributed player, the generation of the node state information structured clustering correction increment operator is based on the nonlinear mapping mechanism of the node state information structured deep collaborative implicit coding vector. Its essence is the dynamic compensation and structural optimization of the linear clustering result. This calculation process relies on the node state information structured deep collaborative implicit coding vector, and through the incremental calculation of the nonlinear activation function and eigenvalue granularity, the deviation of the node state relative to the linear clustering structure is converted into a quantifiable correction parameter. Specifically, a sigmoid function performs a nonlinear transformation on the node state information structured initial linear cluster center encoding vectors, constructing a correlation channel between the global feature skeleton and local dynamic features. The incremental computation of eigenvalue granularity quantifies the nonlinear deviation of node states in the feature space through the interaction between the node state information structured deep collaborative implicit encoding vector and the node state information structured initial linear cluster center activation encoding vector. Dynamic relaxation parameter collaborative resonant coupling optimization further introduces a dynamic compensation mechanism for timing fluctuations and topological correlations to eliminate the interference of environmental noise on the correction increments. Finally, a softmax normalization process is used to generate standardized compensation weights to ensure the fairness and stability of the correction process. The resulting set of node state information structured cluster correction increment operators not only includes compensation parameters for individual nodes but also constructs a global nonlinear compensation strategy through a data-driven adaptive mechanism. Its core value lies in overcoming the inherent shortcomings of linear cluster analysis by capturing the complex distribution patterns of node states beyond the linear cluster structure through deep models, such as the dynamic coupling relationship between network latency and device load, and the implicit influence of regional network topology on synchronization delay. Each incremental operator for structured clustering correction of node state information is based on the initial linear cluster center and enhances the robustness of feature representation through nonlinear correction. While retaining the stability of the global linear structure, it also incorporates nonlinear detail compensation at the micro level. This incremental optimization strategy can adapt to changes in the network environment, such as sudden load changes at edge nodes or cross-regional transmission jitter. It continuously optimizes the parameter weights of the deep model through a backpropagation mechanism, forming a self-learning feature gain reinforcement closed loop. The resulting set of incremental operators for structured clustering correction of node state information not only provides a basis for refined adjustment of node performance evaluation, but also improves the adaptability of the master node election mechanism to complex scenarios through a global nonlinear compensation strategy, ensuring that large-scale clusters can still achieve frame-level precision synchronization under dynamic network conditions.

[0050] In the process of constructing the node state information structured clustering correction increment operator, due to the node state information structured encoding vector x i In addition to the representation of node state information, the node state information structured deep collaborative implicit encoding vector r iThe introduction of multi-scale clustering residual information leads to macroscopic non-stationary imbalance of clustering feature space topology. To this end, the node state information structured encoding vector x is adopted. i The implicit encoding vector r is structured in deep collaboration with node state information i The inner product of <x i ,r i >Construct node status information structured eigenvalue Σ i , and based on this, the structured intrinsic scattering coefficient ∈ i Dynamic relaxation parameter Γ of structured clustering of node state information i :

[0051]

[0052] Where L is the length of the eigenvector, σ(x i -r i ) 2 Represents a vector (x i -r i ) and ∈ i is with Σ i The related node state information structured intrinsic scattering coefficient, the node state information structured clustering dynamic relaxation parameter Γ i A quantitative analysis of the phase relaxation effect of non-integer Hausdorff dimensions for representing multiscale fractal topologies.

[0053] Finally, by establishing a coherent resonance mechanism under energy dissipation constraints, This method effectively solves the stability problem in the nonlinear dynamic evolution of the characteristic space and ultimately forms a complete correction model for the multi-order nonlinear coupling architecture.

[0054] Specifically, in step S1123, based on the set of the node state information structured clustering correction increment operators, the node state information structured linear clustering correction component coding vector of the set of the node state information structured coding vectors is calculated, and the node state information structured linear clustering correction component coding formula is expressed as:

[0055]

[0056] Among them, x bIt is the node state information structured linear clustering correction component encoding vector. It should be understood that the calculation essence of the node state information structured linear clustering correction component encoding vector obtained in this step is the global aggregation of individual node state deviations. By eliminating the noise and redundant information in the node state information structured clustering correction increment operator, the common compensation pattern across nodes is extracted to form a nonlinear correction of the linear cluster center. By mapping the scattered node state deviations into a globally unified node state information structured linear clustering correction component encoding vector, the macro framework of the original linear structure is retained, and the refined correction of nonlinear features is incorporated, thereby constructing a feature benchmark that is more in line with the actual cluster operation state. This process significantly improves the adaptability of the master node election mechanism to dynamic network environments, ensuring that in cross-regional deployment scenarios, even if there are regional network fluctuations or sudden changes in node load, nodes with the best comprehensive performance can still be accurately identified, providing a robust global benchmark reference for frame-level synchronization of large-scale clusters.

[0057] Specifically, in step S1124, the node state information structured linear clustering correction component encoding vector and the node state information structured initial linear clustering center encoding vector are fused to obtain the node state clustering modeling encoding vector, which is expressed as the node state clustering modeling encoding formula:

[0058] v f =α·x b +β·x c

[0059] Among them, α and β are x b and x c The corresponding weighted hyperparameter, v fis the node state cluster modeling encoding vector. It should be understood that through the fusion mechanism, the macroscopic framework of the linear structure and the microscopic correction of nonlinear details can be organically integrated to form a global-local joint representation of the node state distribution. By balancing the complementarity of the linear and nonlinear components, the node state information structured initial linear cluster center encoding vector retains the benchmark characterization of the cluster's global linear structure while incorporating the node state information structured linear cluster correction component encoding vector's ability to fine-tune compensation for individual node nonlinear deviations. This constructs a composite feature representation that combines macroscopic structural stability with microscopic adaptability. This fusion strategy overcomes the representational limitations of a single model, enabling the resulting node state cluster modeling encoding vector to simultaneously reflect the global positioning of node states within the ideal linear distribution framework and nonlinear deviations caused by factors such as network fluctuations and device heterogeneity in actual operating environments. This provides a more comprehensive and accurate decision-making basis for subsequent node performance evaluation and master node election. Thus, after executing this fusion step, the node state cluster modeling encoding vector achieves a multi-dimensional enhancement of the original node state information. On the one hand, the structured skeleton provided by the node state information structured initial linear cluster center encoding vector ensures the global consistency and interpretability of the feature representation. On the other hand, the nonlinear correction information injected by the node state information structured linear cluster correction component encoding vector significantly improves the model's adaptability to complex scenarios. This fusion of features demonstrates greater robustness in dynamic network environments, effectively addressing abnormal scenarios such as sudden changes in node load and regional network fluctuations. It provides a stable and flexible benchmark reference for the master node election mechanism, thereby ensuring that large-scale clusters can still achieve frame-level precision synchronization under complex network conditions, improving the visual expression and user experience of timing-sensitive applications such as light shows.

[0060] Specifically, step S113 is based on the node performance offset value of each node state information structured coding vector in the set of node state information structured coding vectors relative to the node state cluster modeling coding vector to obtain a set of node performance offset values. It should be understood that the node state information structured coding vector describes the specific state of each node, while the node state cluster modeling coding vector represents the optimal operating state of the cluster as a whole. By calculating the node performance offset value, each node can be quantitatively compared with the overall optimal state, clearly showing the degree of difference between each node and the ideal state. This quantitative method makes the evaluation of nodes more objective and accurate, avoiding the shortcomings of evaluating node status based solely on subjective judgment or vague description. In addition, in a complex node cluster environment, different nodes have different adaptability to the current operating environment. By calculating the offset value, the adaptability of each node in the current cluster state can be intuitively reflected, providing data support for the subsequent selection of the master node.

[0061] Specifically, in step S114, the node corresponding to the largest node performance offset value in the set is selected as the master node. It should be understood that the node with the largest offset value indicates the greatest deviation from the overall optimal operating state of the cluster. However, this deviation is not necessarily a disadvantage; rather, it may indicate that the node possesses unique advantages in certain key aspects. For example, it may possess stronger processing power, more stable network connections, or more efficient communication mechanisms, enabling it to stand out in the cluster and assume the special responsibilities of the master node. This differential identification method helps identify nodes that, while deviating from the overall norm, possess unique capabilities to better serve the entire cluster. In the actual operation of large-scale lighting systems, network conditions and device status may change at any time. By selecting the node with the largest offset value as the master node, these changes can be dynamically adapted based on real-time node status information. If the original master node experiences performance degradation or regional network anomalies, the offset values of other nodes may change. In this case, the node with the largest offset value may become a new node that is more suitable for the master node role, thereby enabling automatic master node migration to adapt to the changing environment. This helps improve operational efficiency and avoid synchronization issues or performance bottlenecks caused by improper master node selection.

[0062] In the above-mentioned node synchronization playback method for a distributed player, in step S120, the master node obtains basic information about the video program and calculates playback frame information based on the current playback progress. It should be understood that in the node cluster architecture of a distributed player, the master node bears the core responsibility of coordinating and controlling the synchronized playback of the entire cluster. To effectively perform this responsibility, the master node needs to have basic information about the video program, such as the video's duration, resolution, frame rate, etc., as well as information related to the current playback progress. Only by obtaining this information can the master node accurately perform subsequent calculations and decisions, send correct instructions to the slave nodes, and achieve synchronized playback between the nodes. Specifically, in a specific embodiment of the present application, each node analyzes the video program information added to the node and pre-parses and stores information such as the video format, resolution, frame number, and duration. That is, first, the video information is analyzed to obtain the frame rate and total number of frames of the video program. Then, the playback frame information is calculated in real time based on the current playback progress.

[0063] In the above-mentioned node synchronization playback method for distributed players, in step S130, the master node sends playback status information to the slave node via broadcast. It should be understood that the playback status information includes playback frame information, playback status, basic information of the video program and the current timestamp of the master node. Accordingly, in a distributed node cluster, the number of slave nodes is large and the distribution may be relatively scattered. The use of a broadcast method allows the information sent by the master node to reach all slave nodes at the same time, ensuring that each slave node can receive the latest playback status information, avoiding the situation where information omissions or some nodes cannot receive information due to point-to-point transmission may occur, and ensuring the integrity and consistency of information transmission. In other words, the playback frame information is the key basis for the slave node to determine the current playback screen. The master node sends the playback frame information via broadcast, and the slave node can adjust its own playback progress based on the information and keep in sync with the master node, thereby achieving accurate synchronization of the playback screen between each node and improving the visual effects of scenes such as light shows. Specifically, in a specific embodiment of the present application, real-time playback progress information (status, video program information, video playback progress, current frame information, master node time, etc.) is pushed to other nodes via broadcast. When the master node is in the stop state, the broadcast push time and status information. When the master node is in the play state, the broadcast push time, status, play progress, video program and other information.

[0064] In the above-mentioned node synchronization playback method for a distributed player, in step S140, after receiving the playback status information, the slave node calculates the network delay between the slave node and the master node based on the playback status information and the round-trip time of the data packet. In an embodiment of the present application, the playback status information includes playback frame information, playback status, basic information about the video program, and the current timestamp of the master node. It should be understood that network conditions are complex and changeable, with various factors such as network congestion and signal interference, which can cause data transmission delays. The network connection status between the slave node and the master node is no exception and may change at any time. Therefore, it is necessary to calculate network delay to understand the network status in real time so that the slave node can adjust the playback strategy according to the actual situation and ensure synchronized playback with the master node. The main purpose of calculating network delay is to enable the slave node to accurately adjust the playback according to the delay to achieve precise synchronization with the master node. By obtaining accurate network delay, the slave node can know the time spent on data transmission during the network transmission process and can advance or delay the playback by a corresponding amount to compensate for the delay caused by network transmission, ensure that the content played by the master node is consistent in time, avoid problems such as image desynchronization and audio stuttering, and improve the user experience. In a specific embodiment of the present application, the network delay between the current node and the master node is calculated from the node delay information by the round-trip time point information of the communication packet with the master node. In addition to automatically calculating and processing the network delay, it supports the administrator to customize the adjustment of the delay time to better flexibly ensure the implementation of the synchronization function. Specifically, first, the slave node receives the communication broadcast packet from the master node, records and parses the round-trip time of the data packet to calculate the network delay information between the current node and the master node. Then, the platform is used to customize the configuration delay time (positive / negative, milliseconds) so as to control the synchronization processing of the nodes during the subsequent actual effect joint debugging.

[0065] In the above-mentioned node synchronization playback method for a distributed player, step S150, in which a slave node calculates its target frame to be played based on the playback frame information and the network delay, includes: S151, configuring a custom delay time for the slave node; S152, converting the network delay and the delay time into frame numbers to obtain a first delay frame number and a second delay frame number; and S153, determining the target frame based on the playback frame information combined with the first delay frame number and the second delay frame number. Accordingly, network delay can cause a time lag in the slave node receiving playback status information from the master node, thereby affecting playback synchronization. Furthermore, differences in device performance, hardware clocks, and other factors between different slave nodes can result in time deviations between devices. Configuring a custom delay time can further compensate for these time errors caused by device differences. Converting network delay and custom delay time into frame numbers is done because video playback is based on frames. This allows for a more intuitive translation of time factors into adjustment values related to playback frames, allowing the slave node to accurately determine its target frame to be played, eliminating playback asynchrony issues caused by network and device factors. In this way, by comprehensively considering playback frame information, network latency, and custom delay time, the slave node can accurately calculate the target frame it should currently play, thereby achieving precise frame-level synchronization with the master node and other slave nodes. In scenes such as large-scale light shows, frame-level synchronization can ensure that the images displayed by each node are completely consistent, improving the consistency and expressiveness of the visual effects, and providing the audience with a better viewing experience. Specifically, first, the master node's broadcast packet content information is parsed. Then, the current node's playback progress and frame information are analyzed and obtained. Next, the synchronization frame information is calculated based on the network delay time and the custom delay time. Finally, if the playback progress is delayed or advanced, the playback frame progress is proactively adjusted to synchronize with the master node's current frame progress. Specifically, in one embodiment of the present application, the slave node first parses the progress information pushed by the master node, such as time: 1705304263219 status: playing, program video name: xxxx.mp4, current frame: 1610, etc. Then, the slave node calculates the delay. If the delay is 80ms, the current slave node should play frame 1612. Next, the system checks whether the current node's playback frame rate is within the 1607 / 1617 range. A 5-frame optimization interval is set to avoid frequent synchronization triggering (errors within 10 frames are negligible to the naked eye). When the slave node's progress exceeds this range, synchronization is triggered and the current slave node's playback progress is adjusted to ensure consistency with the master node's progress. Finally, the system loops and waits for the next master node's broadcast packet to be parsed and processed.

[0066] Specifically, the step S151 configures a custom delay time for the slave node. It should be understood that, considering the complexity and uncertainty of the network environment, there may be differences in the network delay between each slave node and the master node. Even in the same network environment, different devices may have different processing delays due to factors such as hardware performance and load. Therefore, in order to ensure that each slave node can accurately follow the playback progress of the master node and avoid the phenomenon of picture asynchrony caused by network jitter or device processing power differences, this application introduces a custom delay time to provide additional flexibility so that the slave node can make more detailed adjustments according to the specific application scenario.

[0067] Specifically, step S152 converts the network delay and the delay time into a number of frames to obtain a first delayed frame number and a second delayed frame number. It should be understood that, considering that during the video playback process, whether it is a master node or a slave node, the content it plays is composed of a series of continuous frames. This means that any time delay will be directly converted into a difference in the number of frames. For example, assuming that the frame rate of the video is 30fps (i.e., 30 frames per second), if a slave node causes a 100 millisecond delay due to network reasons, then this delay is equivalent to 3 frames. If this frame-level delay difference is not taken into account, even a small time delay may cause the problem of picture asynchrony, especially in a large-scale node cluster, such a problem will be magnified, seriously affecting the overall visual effect. Therefore, converting the delay into a number of frames helps to more accurately evaluate and compensate for the impact of these delays, thereby ensuring synchronous playback between nodes. In the technical solution of this application, the first delayed frame number generally refers to the basic delayed frame number calculated based on the network conditions. It is determined based on the round-trip time of the data packet under the current network conditions and reflects the delay situation at the network level. The second delay frame count, on the other hand, involves a more manually set custom delay time, introduced to accommodate specific application requirements or device performance differences. By calculating these two delay frame counts separately, it's possible to make appropriate adjustments based on the specific application environment while maintaining basic synchronization. For example, in situations requiring extremely high synchronization accuracy, such as large-scale light shows, it might be advantageous to increase the second delay frame count to provide more buffer space for potential unexpected situations and ensure good synchronization even during network fluctuations. Furthermore, it's important to note that the process of converting delay to frame counts isn't isolated; it's closely tied to the goal of improving synchronization accuracy. Each slave node must continuously monitor its connection status with the master node and dynamically adjust its playback progress based on the latest delay information. This requires the slave node to not only accurately calculate the frame count corresponding to the network delay and the custom delay, but also to quickly respond and implement appropriate adjustments. For example, if an increase in network delay is detected, the slave node should promptly increase its delay frame count to pre-load the upcoming frame to avoid screen freezes. On the contrary, if the network condition improves and the delay is reduced, the delay frame number should be reduced accordingly to ensure that the playback progress keeps pace with the master node.

[0068] Specifically, the step S153 determines the target frame based on the playback frame information in combination with the first delay frame number and the second delay frame number. It should be understood that by combining the playback frame information with the delay frame number, the target frame that each slave node should play can be calculated more accurately to ensure that they maintain the same playback progress as the master node. Specifically, the present application determines the target frame based on the playback frame information in combination with the first delay frame number and the second delay frame number, which involves an analysis of the current playback frame information and a comprehensive consideration of the delay frame number. For example, at a certain moment, suppose the master node is playing the 1000th frame, and it is known that the frame rate of the video is 30fps (30 frames per second). If a slave node calculates that there is a 50 millisecond network delay (equivalent to 1.5 frames) between it and the master node, and the administrator sets an additional 20 milliseconds custom delay (equivalent to 0.6 frames), the total delay is 70 milliseconds, which is about 2.1 frames. In this case, the slave node needs to add this delay frame number to its current playback frame to determine its target frame. That is, if the slave node is currently playing frame 998, it should adjust its playback progress to jump to frame 1000 plus 2 frames (the delay frame number, rounded up), or start playing at frame 1002, to compensate for network latency and the custom delay. It's important to note that this method of adjusting playback progress based on the delay frame number is not static but needs to be adjusted dynamically based on real-time information. In actual operation, network conditions may change at any time, causing the delay frame number to fluctuate accordingly. Therefore, the slave node needs to continuously monitor the connection quality with the master node and continuously update its target frame number based on the latest delay information. For example, if it detects an increase in network latency, the slave node should promptly increase its delay frame number to preload the upcoming frame content to avoid screen freezes. Conversely, if network conditions improve and latency decreases, the delay frame number should be reduced accordingly to ensure that playback progress closely matches the master node's pace. In addition, we must also consider the situation that when the network conditions are good and no additional custom delay is set, the slave node may be very close to or even ahead of the master node's playback progress. In this case, directly using the delay frame number for adjustment may lead to overcompensation, which will cause the picture to be out of sync. Therefore, when determining the target frame, it is also necessary to incorporate certain optimization strategies, such as setting a reasonable error range, and triggering synchronization adjustment only when the slave node's playback progress deviates from the master node by more than a certain threshold. This not only improves the response speed, but also reduces unnecessary synchronization operations, improving the overall playback smoothness.

[0069] In summary, the node synchronization playback method for a distributed player based on the embodiment of the present application is explained, which performs structured coding and node status clustering analysis on the status information of each node in the node cluster, and then performs node performance offset value to obtain the largest one in the set of node performance offset values, and uses the corresponding node as the master node, and the rest as slave nodes. Then, the master node obtains the basic information of the video, calculates the playback frame information and sends the playback status. Then, the target frame itself is adjusted by the network delay between the slave node and the master node to maintain frame-level synchronization. In this way, even in the face of local network jitter or short-term communication interruption, it can ensure that the nodes operate based on a unified logic, improve the visual coherence and expressiveness in scenes such as light shows, and achieve accurate picture synchronization between large-scale node clusters.

[0070] Figure 6 FIG is a system block diagram of a node synchronization playback system for a distributed player according to an embodiment of the present application. Figure 6 As shown, according to an embodiment of the present application, a node synchronous playback system 100 for a distributed player includes: a node cluster designation module 110, which is used to designate a master node from a node cluster and designate other nodes in the node cluster as slave nodes; a master node playback frame information calculation module 120, which is used to obtain basic information of a video program through the master node and calculate playback frame information based on the current playback progress; a playback status information sending module 130, which is used to send playback status information to the slave node through the master node and using a broadcast method; a network delay calculation module 140, which is used to calculate the network delay between the slave node and the master node based on the playback status information and the round-trip time of the data packet after receiving the playback status information; a current playback target frame calculation module 150, which is used to calculate the target frame that its own slave node should currently play based on the playback frame information and the network delay.

[0071] Here, those skilled in the art will appreciate that the specific operations of each step in the node synchronization playback system for distributed players have been described above with reference to Figures 1 to 5 The description of the node synchronization playback method for distributed players is introduced in detail, and therefore, its repeated description will be omitted.

[0072] As described above, the node-synchronized playback system 100 for a distributed player according to an embodiment of the present application can be implemented in various terminal devices. In one example, the node-synchronized playback system 100 for a distributed player can be integrated into the terminal device as a software module and / or a hardware module. For example, the node-synchronized playback system 100 for a distributed player can be a software module in the operating system of the terminal device, or can be an application developed for the terminal device; of course, the node-synchronized playback system 100 for a distributed player can also be one of the many hardware modules of the terminal device.

[0073] Alternatively, in another example, the node synchronous playback system 100 for distributed players and the terminal device may also be separate devices, and the node synchronous playback system 100 for distributed players may be connected to the terminal device via a wired and / or wireless network and transmit interactive information in accordance with an agreed data format.

[0074] In summary, a node synchronization playback system for a distributed player based on an embodiment of the present application is explained, which performs structured coding and node status clustering analysis on the status information of each node in the node cluster, and then performs a node performance offset value to obtain the largest one in the set of node performance offset values, and uses the corresponding node as the master node, and the rest as slave nodes. Then, the master node obtains the basic information of the video, calculates the playback frame information, and sends the playback status. Then, the target frame itself is adjusted by the network delay between the slave node and the master node to maintain frame-level synchronization. In this way, even in the face of local network jitter or short-term communication interruption, it can ensure that the nodes operate based on a unified logic, improve the visual coherence and expressiveness in scenes such as light shows, and achieve accurate picture synchronization between large-scale node clusters.

Claims

1. A node synchronization playback method for a distributed player, characterized in that: include: Designate a master node from a node cluster and designate other nodes in the node cluster as slave nodes; The master node obtains basic information of the video program and calculates the playback frame information based on the current playback progress; The master node sends playback status information to the slave nodes via broadcast; After receiving the playback status information, the slave node calculates the network delay between the slave node and the master node based on the playback status information and the round-trip time of the data packet; The slave node calculates the target frame that it should currently play based on the playback frame information and the network delay.

2. The node synchronization playback method for a distributed player according to claim 1, characterized in that: Designating a master node from a node cluster and designating other nodes in the node cluster as slave nodes includes: automatically electing a node from the node cluster as the master node based on node status information of each node in the node cluster.

3. The node synchronization playback method for a distributed player according to claim 2, characterized in that: Automatically electing a node from the node cluster as the master node based on node status information of each node in the node cluster, including: Performing structured coding on the node state information of each node to obtain a set of node state information structured coding vectors; Performing an autonomously evolved node state cluster analysis on the set of the node state information structured coding vectors to obtain a node state cluster modeling coding vector; Obtaining a set of node performance offset values based on a node performance offset value of each node state information structured encoding vector in the set of node state information structured encoding vectors relative to the node state cluster modeling encoding vector; The node corresponding to the largest node performance offset value in the set is used as the master node.

4. The node synchronization playback method for a distributed player according to claim 3, characterized in that: Performing an autonomously evolved node state cluster analysis on the set of the node state information structured coding vectors to obtain a node state cluster modeling coding vector, including: Performing linear cluster analysis on the set of node state information structured coding vectors to obtain node state information structured initial linear cluster center coding vectors; Based on the node state information structured initial linear cluster center code vector, calculating the cluster correction increment operator of each node state information structured code vector in the set of node state information structured code vectors to obtain a set of node state information structured cluster correction increment operators; Calculating a node state information structured linear clustering correction component code vector of the set of node state information structured code vectors based on the set of node state information structured clustering correction increment operators; The node state information structured linear clustering correction component encoding vector and the node state information structured initial linear clustering center encoding vector are fused to obtain the node state clustering modeling encoding vector.

5. The node synchronization playback method for a distributed player according to claim 4, characterized in that: Based on the node state information structured initial linear cluster center code vector, calculating the cluster correction increment operator of each node state information structured code vector in the set of node state information structured code vectors to obtain a set of node state information structured cluster correction increment operators, including: Constructing a node state information structured deep collaborative implicit coding vector between each node state information structured coding vector in the set of the node state information structured coding vectors and the node state information structured initial linear clustering center coding vector; Based on the node state information structured deep collaborative implicit coding vector, the characteristic clustering correction incremental operator of each node state information structured coding vector in the set of node state information structured coding vectors relative to the node state information structured initial linear clustering center coding vector is calculated to obtain the set of node state information structured clustering correction incremental operators.

6. The node synchronization playback method for a distributed player according to claim 5, characterized in that: Based on the node state information structured deep collaborative implicit coding vector, calculating the characteristic clustering correction increment operator of each node state information structured coding vector in the set of the node state information structured coding vector relative to the node state information structured initial linear clustering center coding vector to obtain the set of the node state information structured clustering correction increment operators, including: Performing a nonlinear activation process based on a Sigmo id function on the node state information structured initial linear cluster center encoding vector to obtain a node state information structured initial linear cluster center activation encoding vector; Performing incremental calculation based on eigenvalue granularity on the node state information structured initial linear clustering center activation coding vector and the node state information structured deep collaborative implicit coding vector corresponding to the node state information structured coding vector to obtain a node state information structured clustering increment; Performing dynamic relaxation parameter collaborative resonance coupling optimization on the node state information structured clustering increment to obtain the node state information structured clustering optimization increment; The node state information structured clustering optimization increment is normalized based on softmax to obtain a node state information structured clustering correction increment operator corresponding to the node state information structured coding vector.

7. The node synchronization playback method for a distributed player according to claim 1, characterized in that: The play status information includes play frame information, play status, basic information of the video program and the current timestamp of the master node.

8. The node synchronization playback method for a distributed player according to claim 1, characterized in that: The slave node calculates the target frame that it should currently play based on the playback frame information and the network delay, including: Configure the custom delay time of the slave node; Converting the network delay and the delay time into a number of frames to obtain a first delayed frame number and a second delayed frame number; The target frame is determined based on the playback frame information in combination with the first delay frame number and the second delay frame number.

9. A node synchronization playback system for distributed players, characterized in that: include: A node cluster designation module is used to designate a master node from a node cluster and designate other nodes in the node cluster as slave nodes; The master node playback frame information calculation module is used to obtain basic information of the video program through the master node and calculate the playback frame information based on the current playback progress; The playback status information sending module is used to send playback status information to the slave nodes through the master node in a broadcast manner; A network delay calculation module is used to calculate the network delay between the slave node and the master node based on the playback status information and the round-trip time of the data packet after receiving the playback status information; The current playback target frame calculation module is used to calculate the target frame that its own slave node should currently play based on the playback frame information and the network delay.

10. The node synchronization playback system for distributed players according to claim 9, characterized in that: The node cluster designation module is configured to automatically select a node from the node cluster as the master node based on node status information of each node in the node cluster.

Images