Background and Objectives: Long COVID-19 (LC) is a multifaceted condition characterized by persistent fatigue and impaired health-related quality of life (HRQoL). Exercise intolerance and post-exertional symptom exacerbation (PESE) pose challenges for rehabilitation. This study aimed to evaluate the effects of a 12-week core-focused plank exercise program on fatigue and HRQoL in women with LC, using validated patient-reported measures. Materials and Methods: A pilot quasi-experimental design was implemented, with non-randomized group allocation. Thirty-nine women with LC were recruited from the Madrid Long COVID Association. Participants were assigned to either an intervention group (n = 20), which completed a supervised plank-based motor control program, or a control group (n = 19), which maintained usual activity. Fatigue was assessed using the Modified Fatigue Impact Scale (MFIS), and HRQoL was measured using the EQ-5D-5L and EQ Visual Analog Scale (EQ-VAS). Body composition was evaluated via bioelectrical impedance analysis. Results: The intervention group showed significant reductions after intervention in the MFIS total scores compared to the control group, particularly in the physical (21.26 ± 6.76 vs. 25.21 ± 6.06; p < 0.001) and psychosocial domains (4.51 ± 0.41 vs. 5.21 ± 0.38; p < 0.001), without triggering PESE. EQ-VAS scores improved significantly (63.94 ± 15.33 vs. 46.31 ± 14.74; p = 0.034). No significant changes were found in body composition parameters, suggesting that benefits were driven by neuromuscular adaptations rather than morphological changes. Conclusions: A core-focused, non-aerobic exercise program effectively reduced fatigue and improved perceived health status in women with LC. These findings support the use of motor control-based interventions as a safe and feasible strategy for LC rehabilitation, particularly in populations vulnerable to PESE, suggesting clinical applicability for the rehabilitation of women with LC. Further randomized trials are warranted to confirm these results and explore long-term outcomes.

BackgroundNeurological disorders such as stroke, cerebral palsy, Parkinson's disease, and multiple sclerosis frequently cause sensorimotor impairments, limiting independence and quality of life. Pressure garments (PGs), originally designed for burn and vascular conditions, have gained interest in neurorehabilitation for enhancing proprioceptive input and neuromuscular modulation. However, their scope and effectiveness remain unclear.ObjectiveTo map current literature on the application of PGs in neurological disorders and evaluate their effects on sensorimotor function.MethodsA scoping review was conducted following the Arksey and O'Malley framework and PRISMA-ScR guidelines. Five databases and grey literature were searched up to February 2025. Included studies involved PGs used in neurological conditions and reported at least one sensory or motor outcome.ResultsTwenty-three studies were included, covering stroke (n = 7), cerebral palsy (n = 12), multiple sclerosis (n = 3), and Parkinson's disease (n = 1). PGs showed potential benefits in improving proprioception, motor control, and postural stability, especially in stroke and cerebral palsy. However, evidence for spasticity reduction and long-term outcomes was inconsistent. Studies varied in garment type, intervention protocols, and outcome measures, with common methodological limitations.ConclusionPGs may serve as useful adjuncts in neurorehabilitation to enhance sensorimotor function. However, further high-quality studies with standardized protocols are needed to clarify their clinical utility.RegistrationOSF https://doi.org/10.17605/OSF.IO/H9B27.

Background/Objective: Parkinson’s disease (PD) has long been viewed from a neurocentric perspective; however, increasing evidence indicates that glial dysfunction also contributes to dopaminergic neurodegeneration. Although neurotoxic glial phenotypes have been described in amyotrophic lateral sclerosis and Alzheimer’s disease in vivo models, it remains unclear whether similar states arise in the pathological milieu of PD. This study aimed to determine whether glial cells with intrinsic neurotoxic properties emerge in the substantia nigra pars compacta (SNpc) in a PD context. Methods: The classical 6-hydroxydopamine rat model was used to obtain glial cultures from the ipsilateral, toxin-damaged SNpc. These cultures were characterized by quantifying cell number and morphology, as well as by assessing the expression of glial markers. Their neurotoxic potential was evaluated in vitro through co-cultures with PC12 cells, and in vivo by transplanting the isolated cells into the SNpc of naïve rats. Assessments included PC12 cell survival, and integrity of the nigrostriatal pathway and motor performance in the cylinder test. Results: Ipsilateral SNpc cultures yielded 25-fold more cells than contralateral controls. Cultured cells co-expressed astrocytic and microglial markers, thus defining a population of damage-derived reactive glia (DDRG). When co-cultured, DDRG reduced PC12 cell survival, whereas control glial cells showed no neurotoxic effects. In vivo, DDRG transplantation induced a dose-dependent loss of dopaminergic neurons and motor impairments, while vehicle and control glia produced no detectable effects. Conclusions: Our findings suggest that glial cells emerging from a neuroinflammatory/neurodegenerative environment in the SNpc may contribute to dopaminergic neuron loss. Within the context of the experimental PD model used, DDRG appears to represent a glial population with potential pathogenic relevance and may constitute a candidate target for further investigation as a therapeutic strategy in Parkinson’s disease.

Focal hand dystonia (FHD) is a task-specific movement disorder characterized by involuntary muscle contractions during skilled hand use. Impaired intracortical inhibition is a core feature of FHD, yet its relationship to functional brain activity during motor tasks remains poorly understood. This study combined transcranial magnetic stimulation (TMS) and functional MRI (fMRI) to investigate the relationship between intracortical inhibition, as measured by the cortical silent period (cSP), and task-based brain activation in individuals with FHD. Sixteen individuals with FHD and 18 age-matched healthy controls completed cSP assessments and fMRI. Participants performed self-paced finger tapping with each hand separately. For FHD participants with left-hand symptoms (n = 4), left-right labels were flipped during analysis so that right-hand data represented the symptomatic hand and left-hand data represented the non-symptomatic hand across all participants. We analyzed activation in sensorimotor regions and tested voxel-wise correlations between cSP duration and BOLD responses. FHD participants demonstrated greater activation in the inferior parietal lobule (IPL) within the posterior parietal cortex (PPC) during both right- and left-hand finger tapping. Notably, during left-finger tapping (non-symptomatic hand), cSP duration positively correlated with cerebellar activation, suggesting altered integration of inhibitory circuits during motor execution. No such correlations were found during right-finger tapping (symptomatic hand) or for control participants. These findings suggest that individuals with FHD may selectively recruit cerebellar sensorimotor circuits to modulate inhibition during movement of the non-symptomatic hand, though this pattern was not observed during movement of the symptomatic hand. The cerebellum may play a central role in adaptive motor control in FHD. Future work should leverage symptom-inducing tasks and alternative inhibitory markers to further clarify these mechanisms.

Subthalamic nucleus deep brain stimulation (STN-DBS) effectively improves motor symptoms in Parkinson's disease (PD), but long-term survival outcomes and mortality predictors remain unclear. This study aimed to evaluate survival and 10-year clinical outcomes after STN-DBS. We retrospectively analyzed 608 patients with PD who underwent bilateral STN-DBS between 2006 and 2023. Kaplan-Meier and Cox regression analyses assessed survival and preoperative predictors. Motor (MDS-UPDRS Part III), cognitive (MoCA, MMSE), and medication (LEDD) data were compared between baseline and 10 years postoperatively. During a mean follow-up of 4.8 ± 3.5 years, 42 deaths (6.9%) occurred, yielding an age-adjusted mortality rate of 3.3 per 100,000 population. The estimated 5- and 10-year survival rates were 95% and 77%, respectively. The most frequent cause of death was aspiration pneumonia (11.9%). In multivariate Cox analysis, lower preoperative MoCA scores (HR = 0.83, 95% CI 0.73-0.95, p < 0.01) and higher OFF-state MDS-UPDRS Part III scores (HR = 1.03, 95% CI 1.00-1.07, p < 0.05) independently predicted mortality, whereas age at surgery was not significantly associated with survival. Among 118 patients with 10-year follow-up, OFF-state motor and cognitive scores worsened significantly (p < 0.01), while ON-state motor scores (p = 0.21) and total LEDD (p = 0.06) remained stable, suggesting sustained motor control and medication-sparing effects. Long-term survival after STN-DBS in PD was favorable. Preoperative cognitive and motor status, rather than age, determined long-term prognosis, emphasizing the enduring therapeutic value of STN-DBS.

Abstract The selection of electrical stimulation parameters for bio-robots currently relies primarily on empirical trial-and-error rather than neurophysiological principles. This study explored a method for selecting electrical stimulation parameters by analyzing the dynamic characteristics of local field potentials (LFP) in the carp brain motor area during movement.By time-frequency analysis, frequency band relative power calculation, sample entropy and Lempel-Ziv complexity analysis, the response characteristics of carp LFP under static, moving and recovering static state are analyzed systematically. The results show that the relative power of LFP in high frequency band (Theta, Alpha, Beta, Gamma and High Gamma) increases significantly (P&lt;0.05), while that in low frequency band (Delta) decreases significantly (P&lt;0.05), and the signal complexity decreases significantly. Building on these findings, the stimulation frequency significantly affects motor control: optimal performance was achieved at 50–70 Hz, whereas higher or lower frequencies impaired control efficacy. This study establishes a link between LFP features and motor behavior, offering a signal-based parameter selection method to advance precise control of aquatic bio-robots.

The aim of this study was to examine the organization of ball-catching movements with different biomechanical structures, which are among the profile elements in rhythmic gymnastics, based on common synergies. An additional goal was to identify specific synergies presumably formed during the performing of movements with increased coordinative complexity. Synergy parameters were derived from simultaneously recorded electromyographic (EMG) data from 16 muscles and joint angle kinematics; factor analysis and principal component analysis were applied. It was found that at the kinematic level, participants demonstrated similar motor control strategies for catching the ball. Interjoint coordination was structured into two modules, independent of the motor task's complexity. The specificity of kinematic modules manifested in a shift of the main peak of interjoint interaction synchronization, driven by the subjective perception of the ball contact moment. At the muscular level, up to five muscle modules were identified, two of which were consistently present across different ball-catching movements. Their function is associated with generating active motions of the upper limb segments and stabilizing the shoulder girdle. The component composition of muscle synergies is largely determined by the movements' biomechanical structure and the presence of a common subtask within them.

Despite significant progress in acute stroke management, the burden of persistent motor impairments necessitates ongoing research into novel therapeutic strategies. Our study aims to study if antidepressants can effectively improve the motor function and functional independence in patients after stroke. This meta-analysis encompassed Randomized Controlled Trials (RCTs) that enrolled adult stroke patients and compared any antidepressant drug against placebo, and reporting motor outcomes. We excluded studies which reported only cognitive outcomes or single arm studies with no comparator. We searched PubMed, Scopus, Cochrane, and Web of Science for records from inception up to August 2024. Outcomes data were pooled as standardized mean differences (SMDs) with the corresponding 95% confidence intervals (CI). Risk of Bias 2 (RoB2) tool was considered for quality assessment of the included studies. The preliminary search yielded 22 articles. A total of 11,396 patients were included, with a majority being elderly. Anti-depressants, primarily fluoxetine and citalopram, significantly improved Fugl Meyer Motor scale Scores (FMMS) at the endpoint (SMD = 0.79, 95%CI [0.056, 1.02], p-value < 0.00001) and change from baseline (SMD = 0.75, 95%CI [0.27, 1.24], p-value = 0.002), suggesting a substantial positive effect equivalent to a large effect size, potentially reflecting significant improvements in motor control and function. There wasn't a significant subgroup difference between fluoxetine, citalopram, and selegiline. Also, Barthel index (BI) scores endpoints were significantly improved by antidepressants (SMD = 0.54, 95%CI [0.12, 0.96], p-value = 0.01) and change from baseline (SMD = 0.46, 95%CI [0.02, 0.9], p-value = 0.04), indicating a moderate positive effect, likely representing noticeable gains in independence for activities of daily living. There was a significant difference in both BI endpoints score and change from baseline (P < 0.00001) between subgroups favoring escitalopram. However, anti-depressants did not improve modified Rankin Scale (mRS) Scores (SMD = 0.06, 95%CI [-0.01, 0.12], p-value = 0.08). There wasn't a significant subgroup difference between fluoxetine and citalopram (p-value = 0.17). Based on RoB2, 12 studies were rated as having an overall low risk of bias, four were rated as having some concerns, and six were assessed as high risk. Certain antidepressants may enhance motor performance and independence in performing activities of daily living during post-stroke recovery among elderly patients. Fluoxetine was the most common antidepressant described in the literature with significant improvement in motor (FMMS) functional (BI) scales. However, substantial heterogeneity and potential study biases warrant cautious interpretation. Rigorous, large-scale RCTs are necessary to verify these findings and establish long-term safety profiles. They will also help define the optimal therapeutic strategies before routine clinical use is considered.

Symmetry is a core yet controversial concept in the health sciences, spanning multiple disciplines from anatomy and biomechanics to pathophysiology, with important clinical implications for both diagnostic and therapeutic practices [...]

Jemparingan, the traditional seated archery of the Ngayogyakarta Hadiningrat Sultanate, represents a unique cultural heritage that integrates physical practice with spiritual self-development. Addressing the need for character education in the digital age, this study investigates how Jemparingan fosters essential focus and discipline within Generation Z. Utilizing a historical and anthropological approach, the research analyzes the sport’s historical evolution from a royal-exclusive tradition to a modern cultural practice, while examining its philosophical emphasis on the unity of mind and body. The study further evaluates current preservation challenges and identifies key policy efforts necessary to sustain this tradition. Results indicate that the sport’s distinct cross-legged technique demands acute concentration and fine motor control, serving as an effective pedagogical tool for instilling self-regulation. The paper concludes that integrating Jemparingan into modern educational frameworks offers a strategic avenue for cultural continuity and for supporting the psychological development of contemporary adolescents.

Characterizing each person's sensorimotor profile is crucial for designing precise and personalized motor rehabilitation therapies. Building on our previous work in system identification of human motor control dynamics, we now extend our parameter recovery technique developed in synthetic models to a real-world human experiment. This twin-based digital method actively guides the experimental design by selecting the most informative perturbations and movement conditions to most accurately identify (recover) sensory feedback gains. We applied this framework to 10 neurotypical participants, analyzing their performance during arm planar reaching movements. By combining the optimized experimental design with this forward-inverse modeling pipeline, we estimated individual sensory feedback gains. These gains were then used to simulate movement trajectories, achieving a movement prediction accuracy of 85% compared to withheld trajectories performed by the same subjects. These results validate the ability of our mathematical model to capture and explain individual sensorimotor dynamics through the identification of subject-specific feedback gains. This approach offers a promising tool for gaining insights into the roles of different sensory channels and identifying the most informative data required for efficient assessment.

Mental disorders represent a significant challenge for individuals and society. Many of them have a detectable onset during adolescence. Being born preterm or with low birth weight (PTB) has been emerging as a potential risk factor for developing mental health disorders in adolescence. Since PTB infants are considered to be at an increased risk of cognitive and sensory difficulties and are at risk for visual impairments, this systematic review aims to explore (1) whether evidence for a possible interplay between PTB, cognitive, visual, and motor abilities exists in the literature, and (2) whether and how these factors may relate to mental health outcomes of PTB individuals in adolescence. We conducted a registered systematic review following the PRISMA guidelines (PROSPERO #42024513150). The search strategy focused on the databases PubMed, Scopus, PsycINFO and the Cochrane Library and included publications sampling participants born in 1980 or later. Upon screening 499 studies, we analysed 17 studies including a total of 10,842 adolescents aged 11-20 (PTB = 8,813, control = 4,029) published between 2005 and 2022. PTB adolescents exhibited deficits in cognitive and motor domains compared to their full-term peers (≥ 37 gestational weeks; FT), including lower intelligence quotient (IQ), attention and executive function, and motor control. These effects can persist into adolescence and even adulthood. Importantly, several studies demonstrated that PTB adolescents receive diagnoses of psychiatric disorders more often and get diagnosed with more complex psychiatric disorders. Contrarily, evidence for subtle visual alterations and direct links between PTB, the reviewed domains and mental health outcomes remains scarce. These findings highlight that PTB adolescents can face challenges across multiple separate domains, including an elevated prevalence of psychiatric diagnoses. Clarifying the nature of these observational relationships shall provide insights that could improve early detection approaches and targeted intervention in the future.

The evolutionary adaptation of the left inferior frontal gyrus is considered a crucial neural specialization supporting the emergence of human language. As a central node in the language network, it is linked to the temporoparietal cortex via both the ventral and dorsal pathways. These connections enable humans to combine a limited set of vocal elements into infinitely diverse, hierarchically structured sequences. Although homologous brain structures are also present in non-human primates, language remains a uniquely human faculty. This review synthesizes anatomical, functional, and connectivity evidence across species to trace the evolution of the left inferior frontal gyrus in support of language. We argue that language did not emerge from novel cortical areas, but through the gradual repurposing, expansion, and optimization of pre-existing fronto-temporal circuits. Human-specific innovations include vocal neuron specialization, volumetric expansion, strengthened connectivity of the arcuate fasciculus, and a functional shift within the left inferior frontal gyrus from motor control to syntactic processing. Finally, we discuss how lesion studies contribute to our understanding of the brain’s potential for language acquisition and its neurobiological constraints.

Abstract This paper has developed the advanced QIMNSO for enhancement in the DTC of induction motors through a set of proposed schemes on quantum computing, memetic algorithms, adaptiveness of a neural network, and swarm intelligence. The proposed QIMNSO-DTC is designed to avoid main drawbacks inherent in traditional DTC, which are named as torque ripple, response lag, and energy inefficiency-evidently committed under dynamic load conditions. Application of QIMNSO results in prompt torque and flux adjustment, smaller ripples, and reduction of mechanical stress to the motor. This neural network part allows the real-time adaptation of parameters to achieve the best performance for all operating conditions and load variations. Simulation results indicate that QIMNSO-DTC enjoys some merits in comparison to classical control methods, including FOC, SMC, and PID controllers in terms of torque stability, response speed, energy efficiency, and self-adaptiveness. Such enhancements render QIMNSO-DTC very fit for applications where induction motors should be precisely, efficiently, and reliably controlled, such as robotics, electric vehicles, and high-performance industrial drives. QIMNSO represents a promising, scalable control approach in the present research that gives classical methods a large margin of improvement and contributes to further innovation related to intelligent motor control.

Spinal cord injury (SCI) often results in impaired motor control and coordination. Previous studies have highlighted the role of muscle synergies in coordinating motor tasks and their alterations following SCI. However, the adaptation in muscle synergy patterns at the motor unit (MU) level after SCI remains unexplored. This study aimed to investigate MU synergies and clustering in the synergetic soleus and gastrocnemius medialis (GM) muscles and to explore how these patterns are altered in persons with SCI. High-density electromyography (HD-EMG) was used to record MU activity in the soleus and GM muscles of fifteen participants with incomplete SCI and ten non-disabled participants during 20% and 50% maximal voluntary isometric contraction tasks. The HD-EMG signals were decomposed into individual MU spike trains. Inter-muscle coherence analysis was employed to evaluate the shared neural drive between the soleus and GM muscles, and factor analysis was performed to identify synergistic clusters of MUs innervating each muscle. The results showed that both participant groups demonstrated high coherence between the soleus and GM muscles, highlighting a shared neural drive for coordinated function. However, participants with SCI showed altered coherence in the delta frequency band, with significantly higher coherence observed at 50% maximal voluntary contraction (p = 0.047). Additionally, factor analysis revealed that participants with SCI had a reduced proportion of MUs in the shared cluster within the GM muscle at 20% maximal voluntary contraction (p < 0.01). These findings suggested that SCI may disrupt MU synergies and clustering, potentially impairing motor coordination. This research offered valuable insights into the underlying mechanism of muscle synergies and the neural adaptations following SCI, providing crucial information for the development of future rehabilitation strategies.

Transcranial magnetic stimulation (TMS), a non-invasive neuromodulation technique, is widely employed in treating various neurological and psychiatric disorders due to its favorable safety and tolerability profile. While traditionally recognized for its role in motor control, accumulating evidence implicates the cerebellum in regulating non-motor functions, including cognition. Historically, TMS research predominantly targeted cortical brain regions. However, leveraging the functional and structural characteristics of the cerebellum, recent investigations have increasingly focused on the cerebellum as a stimulation target, exploring the effects of cerebellar transcranial magnetic stimulation (CRB-TMS) on cognitive function.This review presents an overview of the current understanding and status of the cerebellar-cognitive connection. It critically examines the effects of cerebellar CRB-TMS on different cognitive functions, such as memory, language, attention and executive function .And it discusses the effects of different stimulation modes and stimulation sites on cognitive functions. Furthermore, it explores the research progress and potential clinical applications of CRB-TMS for cognitive disorders. We summarize key findings, discuss underlying mechanisms, and outline future research directions to inform the optimization of CRB-TMS for modulating cognitive function.

There is growing recognition that short-term changes in speech perception influence speech production. These effects offer new insight into interactions of perception and production and shed light on phonetic convergence, the subtle alignment of speech patterns that emerges between communication partners. Across three experiments, we investigate the representations underlying perceptual effects on speech production. Building from the established influence of preceding context on speech perception, we strategically pair contexts to shift perception of target syllables and test whether these perceptual effects influence speech production. Experiment 1 shows that speech contexts rich in articulatory-phonetic information shift speech perception and alter acoustic patterns of speech production. Experiment 2 demonstrates that continuous natural speech filtered to possess subtly different spectral profiles that do not impact articulatory-phonetic information also affects both perception and production. Strikingly, Experiment 3 reveals that even nonspeech tones induce perceptual context effects that influence speech production. The findings point to a much broader scope of perception-production transfer than reported previously and challenge the necessity of social interaction, covert imitation, and articulatory-phonetic information in sensorimotor speech interactions. This emphasizes the need to extend models of speech motor control to account for perceptual influences of other talkers' speech on speech production and to accommodate general auditory processes in sensorimotor models of speech. (PsycInfo Database Record (c) 2026 APA, all rights reserved).

Flexible motor control is essential for navigating complex, unpredictable environments. Although movement execution is often associated with stereotyped patterns of neural and muscular activation, the degree to which these patterns are conserved versus flexibly reorganized to meet task demands across diverse contextual changes has not been well characterized. Here we recorded head and body kinematics alongside muscle activity in rhesus monkeys during head stabilization-crucial for maintaining gaze and balance-while walking on a treadmill at various speeds, and during overground locomotion in the presence or absence of enhanced autonomic arousal. Dimensionality reduction analyses revealed a flexible control strategy during treadmill walking: a stable activation structure that scaled with speed. In contrast, overground walking evoked heightened muscle engagement and more substantial changes in organization. This pattern largely persisted even during elevated arousal, with larger pupil size linked to stronger but structurally preserved muscle recruitment. Together these findings demonstrate that the brain dynamically adapts motor coordination to context even for automatic behaviors, underscoring the need to examine control strategies in a wide range of conditions.

Focal hand dystonia (FHD) severely impairs task-specific motor control, yet the optimal surgical target for stereotactic intervention remains uncertain. This study aimed to identify the precise thalamic lesion site associated with symptomatic improvement and to clarify its network connectivity. We retrospectively analyzed 164 patients with FHD (mean age = 42.0 years, 26.2% women) who underwent stereotactic thalamotomy of the ventral lateral thalamus. Voxel-wise probabilistic lesion mapping was applied to relate lesion locations to clinical outcomes. Structural connectivity analyses assessed fiber tracts linked to the optimal lesion site. Model performance was evaluated by 10-fold cross-validation, validation in an out-of-sample cohort, and testing in reoperation cases. We identified that lesioning the border zone between the ventralis oralis posterior (Vop) and ventralis intermedius (Vim) nuclei was associated with improvement of FHD. The predictive model achieved high accuracy in cross-validation (area under the curve [AUC] = 0.836) and performed robustly in independent validation. Connectivity analyses showed that the Vop-Vim border zone was linked to cerebellothalamic and pallidothalamic afferents as well as thalamocortical projections to the supplementary motor area and premotor cortex. In contrast, lesions extending into the ventralis oralis anterior nucleus were associated with an increased risk of motor complications. Precise targeting of the Vop-Vim border maximizes clinical benefit while minimizing adverse effects in FHD thalamotomy. These findings establish the first evidence-based thalamic target for FHD, offering practical guidance for stereotactic interventions and advancing understanding of dystonia pathophysiology. ANN NEUROL 2026.

Abstract Accurate measurement of key parameters in permanent magnet synchronous motor (PMSM) is of paramount importance for motor control and monitoring. To reduce costs and overcome the limitations of sensor accuracy, there has been increasing attention and research on establishing key parameter prediction models using easily accessible data. However, the existing soft sensor models are established based on the measured data, neglecting the known mechanism knowledge. To this end, an improved soft sensor modelling framework combining the virtual output of the mechanism model is proposed to enhance the prediction accuracy. In this work, the mechanistic model of PMSM is briefly introduced. The fusion model, using the Transformer as an example, is denoted as F-Transformer. Then F-Transformer is established based on the easily measured variables and the virtual output of the mechanism model. The modelling performance and advantages of the proposed model are verified by a semi-physical experimental platform. The average root mean square error of the F-Transformer was reduced by approximately 87.73% and 88.51% compared to the mechanism model and data-driven model(Transformer) respectively, fully demonstrating its advantages in high prediction accuracy and strong generalization capability under various operating conditions.