Data Compression Research Articles

EFFICIENCY OF LOSSLESS DATA COMPRESSION ALGORITHM MODIFICATION

The current level of development of information technologies causes a rapid increase in the amount of information stored, transmitted and processed in computer systems. Ensuring the full and effective use of this information requires the use of the latest improved algorithms for compaction and optimization of its storage. The further growth of the technical level of hardware and software is closely related to the problems of lack of memory for storage, which also actualizes the task of effective data compression. Improved compression algorithms allow more efficient use of storage resources and reduce data transfer time over the network. Every year, programmers, scientists, and researchers look for ways to improve existing algorithms, as well as invent new ones, because every algorithm, even if it is simple, has its potential for improvement. A wide range of technologies related to the collection, processing, storage and transmission of information are largely oriented towards the development of systems in which graphical presentation of information has an advantage over other types of presentation. The development of modern computer systems and networks has influenced the wide distribution of tools operating with digital images. It is clear that storing and transferring a large number of images in their original, unprocessed form is a rather resource-intensive task. In turn, modern multimedia systems have gained considerable popularity thanks, first of all, to effective means of compressing graphic information. Image compression is a key factor in improving the efficiency of data transfer and the use of computing resources. The work is devoted to the study of the modification of the data compression algorithm The Quite OK Image Format, or QOI, which is optimized for speed for the compression of graphic information. Testing of those implementations of the algorithm, which were proposed by its author, shows such encouraging results that it can make it competitive with the already known PNG algorithm, providing a higher compression speed and targeting work with archives. The article compares the results of the two proposed modifications of the algorithm with the original implementation and shows their advantages. The effectiveness of the modifications and the features of their application for various cases were evaluated. A comparison of file compression coefficients, which were compressed by the original QOI algorithm, with such coefficients, which were obtained as a result of the application of modifications of its initial version, was also carried out.

Architectural Design of Representational State Transfer Application Programming Interface with Application-Level Base64-Encoding and Zlib Data Compression

Representational State Transfer (REST) is an architectural style that underlies the protocol of HyperText Transfer Protocol (HTTP)and is used by web applications. The implementation of REST principles in web services is often known as RESTful API. The standard communication used in RESTful APIs uses HTTP without involving data compression. There are quite a lot of RESTful API clients in the form of mobile applications that use metered networks. Thus, reducing the required bandwidth during transmission is suggested to be beneficial. The data compression technique is widely known to reduce data size, but this additional step may yield side effects such as increased memory usage and processing time. This study aims to investigate how the data compression process, which specifically uses Zlib and Base64 encoding, may put additional load on the whole process of delivering content to a RESTful API. The performance and characteristics in terms of bandwidth saved when distributing JSON data from a RESTful API in compressed format are also be investigated. According to the experiment result, it is suggested that the compression process can reduce network bandwidth by up to 66% with negligible additional memory usage for the compression and decompression processes.

Compression Techniques for Real-Time Control and Non-Time-Critical Big Data in Smart Grids: A Review

Significant amounts of data need to be transferred in order to optimize the operation of power grids. The development of advanced metering and control infrastructure ensures a growth in the amount of data transferred within smart grids. Data compression is a strategy to reduce the burden. This paper presents current challenges in the field of time-series data compression. This paper’s novel contribution is the division of data in smart grids to real-time data used for control purposes and big data sets used for non-time-critical analysis of the system. Both of these two applications have different requirements for effective compression. Currently used algorithms are listed and described with their advantages and drawbacks for both of these applications. Details needed for the implementation of an algorithm were also provided. Comprehensive analysis and comparison are intended to facilitate the design of a data compression method tailored for a particular application. An important contribution is the description of the influence of data compression methods on cybersecurity, which is one of the major concerns in modern power grids. Future work includes the development of adaptive compression methods based on artificial intelligence, especially machine learning and quantum computing. This review will offer a solid foundation for the research and design of data compression methods.

PRINCIPAL COMPONENTS VERSUS AUTOENCODERS FOR DIMENSIONALITY REDUCTION: A CASE OF SUPER-RESOLVED OUTPUTS FROM PRISMA HYPERSPECTRAL MISSION DATA

Abstract. This study attempts to solve these issues associated with hyperspectral (HS) data, i.e., coarse spatial resolution and high volume, by understanding the effect of deep learning and traditional dimensionality reduction on super-resolved products generated from the recently launched PRecursore IperSpettrale della Missione Applicativa (PRISMA) HS mission. Four single-frame super-resolution (SR) algorithms have been used to super-resolve a 30 m PRISMA scene of Ahmedabad, India and generate 15 m spatial resolution images with both spatial and spectral fidelity. Iterative back projection (IBP) and sparse representation (SIS) are the best and worst-performing SR algorithms following a comparative assessment and validation protocol. Next, denoising autoencoders and PCT computed using singular and eigenvalue decompositions have been executed on the original PRISMA, IBP and SIS-based super-resolved datasets. The resulting low-dimensional representations have been assessed to preserve the original dataset's topology using label-independent Lee and Verleysen's co-ranking matrix and loss of quality measure. Findings suggest that autoencoders are computationally expensive and require a higher neighbourhood size than PCT and its variants to produce a high-quality encoding. These insights remain significant for urban information extraction as there are few direct comparative assessments between machine learning-based linear and non-linear data compression methods in earlier studies.

Performance analysis of IoT networks in terrestrial environment utilizing LZW data compression technique

The proliferation of Internet of Things (IoT) devices in terrestrial environments necessitates efficient data transmission and management protocols to optimize network performance. This research paper presents a comprehensive study on the implementation of IoT networks in a terrestrial setting, focusing on the integration of LZW (Lempel-Ziv-Welch) compression with three prominent IoT protocols: CoAP (Constrained Application Protocol), M2M (machine-to-machine), and MQTT (Message Queuing Telemetry Transport). We did a thorough test of these protocols using the NetSim simulator. We looked at how well they worked in terms of throughput, en-to-end delay, and routing overhead, all of which are important for loT network efficiency. The point of our study is to find out what the pros and cons of these protocols are when they are combined with LZW compression in a terrestrial loT setting. The results of our analysis reveal significant variations in the performance of CoAP, M2M, and MQTT when subjected to data compression. Throughput efficiency, which directly impacts data transmission rates, is scrutinized alongside end-to-end delay, a critical factor in ensuring timely data delivery. Additionally, we explore the implications of protocol choice on routing overhead, a crucial metric for network resource utilization. This research contributes to the ongoing efforts to optimize IoT networks in terrestrial settings, offering valuable guidance to network architects and developers seeking to strike the right balance between protocol selection and data compression for improved IoT performance.

Discrete Wavelet Analysis: A Mighty Approach for Image Segmentation

This paper explores the application of Discrete Wavelet Analysis, a mathematical and signal processing technique, in the context of image segmentation, which provide a pixel-level or region-level decomposition of the image, enabling the extraction of relevant information for subsequent analysis and interpretation. Introducing the basic image segmentation techniques and the DWA, this paper discovers that DWA has found widespread application in fields such as signal processing, image analysis, and data compression. Compared with Fourier Transform, DWA is more suitable for image segmentation, having unique advantages and characteristics. Among the procedures of image segmentation, the most important point is feature selection, which determine the criteria for distinguishing different regions within the image. Despite DWA has many advantages, this technology also owns many challenges and limitations, which may be solved by lasting academic research to refine and extend Discrete Wavelet Analysis methodologies for image segmentation. In short, this research highlights the promise of Discrete Wavelet Analysis, emphasizing the use of high- quality image processing.

Flexible and accurate total variation and cascaded denoisers-based image reconstruction algorithm for hyperspectrally compressed ultrafast photography

Hyperspectrally compressed ultrafast photography (HCUP) based on compressed sensing and time- and spectrum-to-space mappings can simultaneously realize the temporal and spectral imaging of non-repeatable or difficult-to-repeat transient events with a passive manner in single exposure. HCUP possesses an incredibly high frame rate of tens of trillions of frames per second and a sequence depth of several hundred, and therefore plays a revolutionary role in single-shot ultrafast optical imaging. However, due to ultra-high data compression ratios induced by the extremely large sequence depth, as well as limited fidelities of traditional algorithms over the image reconstruction process, HCUP suffers from a poor image reconstruction quality and fails to capture fine structures in complex transient scenes. To overcome these restrictions, we report a flexible image reconstruction algorithm based on a total variation (TV) and cascaded denoisers (CD) for HCUP, named the TV-CD algorithm. The TV-CD algorithm applies the TV denoising model cascaded with several advanced deep learning-based denoising models in the iterative plug-and-play alternating direction method of multipliers framework, which not only preserves the image smoothness with TV, but also obtains more priori with CD. Therefore, it solves the common sparsity representation problem in local similarity and motion compensation. Both the simulation and experimental results show that the proposed TV-CD algorithm can effectively improve the image reconstruction accuracy and quality of HCUP, and may further promote the practical applications of HCUP in capturing high-dimensional complex physical, chemical and biological ultrafast dynamic scenes.

A Real-Time and Robust Neural Network Model for Low-Measurement-Rate Compressed-Sensing Image Reconstruction.

Compressed sensing (CS) is a popular data compression theory for many computer vision tasks, but the high reconstruction complexity for images prevents it from being used in many real-world applications. Existing end-to-end learning methods achieved real time sensing but lack theory guarantee for robust reconstruction results. This paper proposes a neural network called RootsNet, which integrates the CS mechanism into the network to prevent error propagation. So, RootsNet knows what will happen if some modules in the network go wrong. It also implements real-time and successfully reconstructed extremely low measurement rates that are impossible for traditional optimization-theory-based methods. For qualitative validation, RootsNet is implemented in two real-world measurement applications, i.e., a near-field microwave imaging system and a pipeline inspection system, where RootsNet easily saves 60% more measurement time and 95% more data compared with the state-of-the-art optimization-theory-based reconstruction methods. Without losing generality, comprehensive experiments are performed on general datasets, including evaluating the key components in RootsNet, the reconstruction uncertainty, quality, and efficiency. RootsNet has the best uncertainty performance and efficiency, and achieves the best reconstruction quality under super low-measurement rates.

Audio data compression affects acoustic indices and reduces detections of birds by human listening and automated recognisers

ABSTRACT Increasing popularity in passive acoustic monitoring and the ease with which researchers can accumulate large quantities of acoustic data has resulted in challenges for audio recording storage, archiving, and management. Reductions in file size can be achieved by lowering sample rate and compressing to different formats; however, how these processes affect audio data quality, and the resulting interpretation of wildlife data is not well understood. We investigated the effect of sampling rate and lossy compression of audio recordings to MP3 from their native WAV format on the performance of four commonly applied avian bioacoustic applications: community listening, distance estimation, automated recognition, and acoustic indices. Compression to MP3 decreased the number of detections, including a reduction in total abundance of individuals when transcribing audio files for community listening and lower precision and recall for automated recognisers. Sampling rate reduction introduced systematic bias to acoustic indices and had an influence on precision and recall for recognisers as well. We recommend against the use of MP3 compression to reduce file volume and suggest other lossless forms of audio compression where an exact copy of the original recording can be recovered.

Interference fading suppression for distributed acoustic sensor using frequency-shifted delay loop

In order to suppress the interference fading of phase-sensitive optical time domain reflectometer (Φ-OTDR), a multi-frequency detection method based on frequency-shifted delay loop (FSDL) is proposed. A probe pulse train with up to five consistent frequency components are generated by periodic frequency-shifted, amplification, delay and filtering processing in FSDL. The piecewise aggregate approximation (PAA) is introduced to compress the amount of operational data. Wavelet energy spectrum analysis and support vector machine (SVM) are used to extract features and realize the fading classification. With the help of fading label making and restoration, the multi-frequency signals are intelligently aggregated using rotated-vector-sum (RVS). Experimental results show that PAA has applicability in data compression of beat signals. After 5-layer wavelet energy spectrum analysis, the fading binary classification accuracy can reach 96.77 %, and the fading signals identified after segmentation can be restored to the original data. The fading probability based on SVM output labels can be reduced to 1.89 %. The SVM-based classification output labels aggregation shows that the vibration positioning signal-to-noise ratio (SNR) can be increased to 13.52 dB, the phase demodulation curve is smoother, and the demodulation SNR can be up to 17.63 dB. Finally, the R-Square of 0.997 for the amplitude of demodulated vibration signal in repeatability experiment verifies the validity of this effective scheme for fading recognition and suppression.

A novel cell phone localization solution for trapped victims based on compressed RSSI fluctuation range and PSO-BP neural network

With the frequent occurrence of global earthquake disasters in recent years, quickly and accurately locating trapped people is the key to reducing the mortality rate of earthquake disasters. This paper proposes a cell phone localization solution for trapped people that can be rapidly deployed after an earthquake. This solution locates the buried cell phone by collecting its received WiFi signal strength indication (RSSI) data. Reducing the fluctuation of RSSI value is crucial to accurately locating the buried cell phone. We filter RSSI data outliers by hybrid filtering and propose a fluctuation range compression factor Q to reduce the fluctuation range of RSSI data. In order to adapt to the compressed RSSI data fluctuation range, the processed RSSI data were grouped. The grouped data are then imported into the PSO-BP neural network training to build the ranging model. So that there is a mapping relationship between RSSI data and trapped distance. The established ranging model is denoted as the HFQBP ranging model. After establishing the ranging model, the cell phone is localized by an Optimized Multilateration (OM) algorithm. For simplicity, this algorithm is named HFQBP-OM. In the experimental phase, we combined the method proposed in this paper with other methods, resulting in three different localization algorithms: HFQBP-WSC, PL-OM, and PL-WSC. And compared with the algorithms in this paper. The results indicate that the algorithm HFQBP-OM proposed in this paper has improved the total average accuracy of the nine localization points by 34.7%, 54.8%, and 58.9% compared to HFQBP-WSC, PL-OM, and PL-WSC, respectively. The HFQBP-OM algorithm has higher accuracy and stability. Also, it should be noted that the localization solution proposed in this paper uses smartphones connected to WiFi probes to collect RSSI data of the WiFi signal of the buried cell phone. Use a laptop computer to solve the localization coordinates. Smartphones and laptops are so popular that WiFi probes are the only special equipment needed. WiFi probe is inexpensive. It reduces the difficulty of implementing the localization solution in this paper and increases the generalizability of the solution.

Homomorphic Compression: Making Text Processing on Compression Unlimited

Lossless data compression is an effective way to handle the huge transmission and storage overhead of massive text data. Its utility is even more significant today when data volumes are skyrocketing. The concept of operating on compressed data infuses new blood into efficient text management by enabling mainly access-oriented text processing tasks to be done directly on compressed data without decompression. Facing limitations of the existing compressed text processing schemes such as limited types of operations supported, low efficiency, and high space occupation, we address these problems by proposing a homomorphic compression theory. It enables the generalization and characterization of algorithms with compression processing capabilities. On this basis, we develop HOCO, an efficient text data management engine that supports a variety of processing tasks on compressed text. We select three representative compression schemes and implement them combined with homomorphism in HOCO. HOCO supports the extension of homomorphic compression schemes through a modular and object-oriented design and has convenient interfaces for text processing tasks. We evaluate HOCO on six real-world datasets. The three schemes implemented in HOCO show trade-offs in terms of compression ratio, supported operation types, and efficiency. Experiments also show that HOCO can achieve higher throughput in random access and modification operations (averagely 9.18× than the state-of-the-art) and lower latency in text analytic tasks (averagely 7.16× than processing on uncompressed text) without compromising compression efficacy.

High-Ratio Compression for Machine-Generated Data

Machine-generated data is rapidly growing and poses challenges for data-intensive systems, especially as the growth of data outpaces the growth of storage space. To cope with the storage issue, compression plays a critical role in storage engines, particularly for data-intensive applications, where a high compression ratio and efficient random access are essential. However, existing compression techniques tend to focus on general-purpose and data block approaches, but overlook the inherent structure of machine-generated data and hence result in low compression ratios or limited lookup efficiency. To address these limitations, we introduce the Pattern-Based Compression (PBC) algorithm, which specifically targets patterns in machine-generated data to achieve Pareto-optimality in most cases. Unlike traditional data block-based methods, PBC compresses data on a per-record basis, facilitating rapid random access. Our experimental evaluation demonstrates that PBC, on average, achieves a compression ratio twice as high as the state-of-the-art techniques while maintaining competitive compression and decompression speeds. We also integrate PBC to a production database system and achieve improvements on both comparison ratio and throughput.

Uncertainty analysis of UAM TMI-1 benchmark by STREAM/RAST-K

This study rigorously examined uncertainty in the TMI-1 benchmark within the Uncertainty Analysis in Modeling (UAM) benchmark suite using the STREAM/RAST-K two-step method. It presents two pivotal advancements in computational techniques: (1) Development of an uncertainty quantification (UQ) module and a specialized library for the pin-based pointwise energy slowing-down method (PSM), and (2) Application of Principal Component Analysis (PCA) for UQ. To evaluate the new computational framework, we conducted verification tests using SCALE 6.2.2. Results demonstrated that STREAM's performance closely matched SCALE 6.2.2, with a negligible uncertainty discrepancy of ±0.0078% in TMI-1 pin cell calculations. To assess the reliability of the PSM covariance library, we performed verification tests, comparing calculations with Calvik's two-term rational approximation (EQ 2-term) covariance library. These calculations included both pin-based and fuel assembly (FA-wise) computations, encompassing hot zero-power and hot full-power operational conditions. The uncertainties calculated using both the EQ 2-term and PSM resonance treatments were consistent, showing a deviation within ±0.054%. Additionally, the data compression process yielded compression ratios of 88.210% and 92.926% for on-the-fly and data-saving approaches, respectively, in TMI fuel assembly calculations. In summary, this study provides a comprehensive explanation of the PCA process used for UQ calculations and offers valuable insights into the robustness and reliability of newly developed computational methods, supported by rigorous verification tests.

ChainedFilter: Combining Membership Filters by Chain Rule

Membership (membership query/membership testing) is a fundamental problem across databases, networks and security. However, previous research has primarily focused on either approximate solutions, such as Bloom Filters, or exact methods, like perfect hashing and dictionaries, without attempting to develop an integral theory. In this paper, we propose a unified and complete theory, namely chain rule, for general membership problems, which encompasses both approximate and exact membership as extreme cases. Building upon the chain rule, we introduce a straightforward yet versatile algorithm framework, namely ChainedFilter, to combine different elementary filters without losing information. Our evaluation results demonstrate that ChainedFilter improves performance of many applications including static dictionary, lossless data compression, Cuckoo Hashing, LSM-Tree and Learned Filters.

Fault detection using generalized autoencoder with neighborhood restriction for electrical drive systems of high-speed trains

Over the past two decades, fault detection of high-speed trains has become an active issue in the transportation area. Recent work has demonstrated the benefits of autoencoder for fault detection in high-speed trains. The latent variable of autoencoder is generally viewed as features generated by the learning process, without explainable interpretation and assessment of the latent variable with respect to its information quality. In the context of representation learning, the latent variable of autoencoder plays an important role towards an fault detection. There is a lack of a methodical framework for explainable applications of AE-based technique to approach fault detection issues. This paper proposes a generalized autoencoder based on neighborhood restriction (NRGAE) to minimize the information loss caused by data compression and reconstruction. The main objective of this paper is to establish a framework for learning autoencoder-based fault detection in the electrical drive systems of high-speed trains. Meanwhile, this work introduces the concepts of mutual information and local linear embedding for electrical drive systems and fault detection specifications. A case study on an electrical drive system is given to illustrate the capability of proposed fault detection.

DUAL-C: Building a “soft error efficient” on-the-fly compression mechanism for raw video data at edge devices

Nowadays, various edge applications play an increasingly important role in our life, in which video data accounts for a dominant portion of the data traffic. Although standard compression methods, such as H.264 and MPEG4, can considerably reduce the wireless bandwidth for transmission energy savings of edge devices, the large volume of raw video streaming data is still buffered in edge device video buffer, constituting non-negligible energy consumption. Beyond energy issues, edge video data has been increasingly applied in safety–critical scenarios, where reliability has also become a critical concern. Unlike the controlled accuracy sacrifice of the lossy compression mechanism, the high information content per bit may render compressed data more sensitive to random transient faults attacking (i.e., soft errors), thereby inducing severe safety hazards in some video processing tasks (e.g., autonomous vehicle collision).Thus, in this paper, we propose a “soft error efficient” on-the-fly compression mechanism named DUAL-C to achieve energy savings in edge device video buffer while maintaining data reliability. Firstly, observing the duplicate feature of upper bits and approximable characteristic of lower bits within the camera sensors generated streaming raw pixel blocks, we design two compressing techniques: DU-C extracts the common duplicate upper-bits while AL-C merely stores one representative approximate lower-bits for all pixels within a block. By combining DU-C and AL-C with the Sampling-based Approximation paradigm (SA), DUAL-C achieves the overall energy consumption reduction of the video buffer by 71.31%. Secondly, to avoid the reliability “dark side” introduced by compression, we conduct a comprehensive error resilience analysis of compressed data and merely harden the error-critical control bits and common upper bits with ECC. Evaluation results show that DUAL-C compressed data exhibits even better error robustness than its uncompressed counterpart with negligible protection overhead.

UNDAMENTALS OF INFORMATION AND COMPUTATION IN THE REALM OF THE QUANTA

Information is a physical. Physical systems register and process information. These facts generated enormus interest in the development of novel quantum tecnologies, especially because the construction of smaller electronic devices ultimately leads to a consideration of quantum mechanical effects in electronic and computer desingns. The notion of the classical bit of information theory was formally pushed into the realm of the quanta with the introduction of the quantum bit or qubit, in the seminal works of Deutsch (Deutsch, 1985) and Shor (Shor, 1994). They demostrated that, indeed, controlled multipartite qubit interference effects could provide the means for a radical new way of computing, allowing the computation of many intractable computational problems, such as the factoring of large number or exact simulation of large quantum systems. The field of experimental and theoretical research in quantum information and computation has emerged as very important player in the understanding of quantum phenomena the both the basic and technological levels. This has attracted the attention of numerous researchers with backgrounds ranging from computer science, mathematics and engineerin, to the physical sciences, and we now have an interdisciplinary field where great efforts are being made in order to build devices that allow the progressing of information at a quantum level.
 A concise introduction to the field of quantum information and quantum computation is presented. This stars with the bassic definitions of bits, quantum registers, through to the universal gate-set for building the universal quantum computer, from a quantum network model of computation. The work shows how two-qubit gates suffice for quantum computation, emphasing the power of the quantum circuit representation for entagling and disentagling quantum states. This leads to the “no-cloning theorem”, which leads us to many interesting applications, such as quantum cryptography. Two alternative approaches or performing quantum computation are also described i) the one-way or measurement based quantum computer method, and ii) holonomic or geometric quantum computation. Following this, quantum entaglement quantification is highlighted, particulary its usefulness as a communication resource, in order to describe some of its most celebrated practical applications to date: quantum teleportation, cryptography, dens coding, and data compression. Deutsch’s concept of quantum parallelism is enphasized in order to gain insight into the potential for efficiently solving certain classically intratable algorithms. A subject central to the field of QIP- quantum decoherence- is then introduced.. Possible ways to overcome it, in particular quantum error correction, are discussed. A descriptios of some of the currently available hardware of the practical implementation of quantum computation is provided with a discussion of the main physical quantum bits that are currently employed (or proposed) for a such a purpose.

Reliability-Aware Task Processing and Offloading for Data-Intensive Applications in Edge computing

With the growing popularity of mobile edge computing (MEC), a number of data-intensive applications have been deployed. Quality of Service (QoS), as one of the most important requirements in MEC, has attracted significant attention of both academia and industry. There are two challenges for ensuring the Qos of the data-intensive services. Firstly, besides performance, reliability is another important concern especially for some critical applications, which remains unexplored. Secondly, data compression has to be involved for reducing the heavy workload of wireless communications at the edge of the network. To overcome these challenges, this paper jointly studies the reliability-aware data compression and task offloading for data-intensive applications in MEC. Markov models are constructed to capture the dynamics of the state transitions of MEC systems, and quantitative analyses are conducted for performance and reliability evaluation. For fully taking advantages of data compression in QoS guarantee, we formulate an optimization problem with the objective of minimizing latency while satisfying constraints on reliability and energy consumption. After describing the NP-Hardness of the problem, we apply the techniques of problem linearization and constraint relaxation, transform the original problem into a convex optimization problem. Then, the problem can be solved in a parallel way by introducing the Alternating Direction Multiplier Method (ADMM), and a distributed Reliability-aware Task Processing and Offloading (RTPO) algorithm is presented. Finally, extensive simulation experiments are conducted to validate the efficacy of our approach, the experimental results illustrate the superiority of our approach over both baseline and state-of-the-art algorithm.

Latent ODEs-based data compression method for power system

In order to effectively compress the power system data and reduce the space occupied by data storage and transmission, this paper proposes a data compression method based on Latent Segmented ODE (LatSegODE) for the large-scale steady-state data generated during the working process of the power system. By training the ODE model using a large amount of steady-state data, a latent ODE model is generated to characterize the steady-state data of the power system, thereby significantly reducing the storage space occupied by the data. During the experimental stage, compression testing is carried out on the static data of the power system of a power grid company. According to the quantitative evaluation results of the two indices of normalized mean square error and data compression ratio, the proposed method has obvious compression advantages, and the compressed signals have a high degree of fitting to the actual sampled signal waveforms.