Abstract: Abstract: In this paper, the research progress of the camellia oleifera fruit harvesting equipment was introduced. This paper simplified the structure of the 2P4R hybrid camellia oleifera fruit harvesting robot, which included waist part, arm part and wrist part. The manipulator could accomplish 6 kinds of movements including waist revolution, translational motion of vertical slider and horizontal slider, and 3 kinds of revolute motions of the wrist part. In arm part, the fore-arm is linked with back-arm by 2 components; one is link-bar below and the other is lower-arm above. Fore-arm is paralleled with back-arm and link-bar is paralleled with lower-arm. It means these 4 components form a parallel quadrilateral mechanism, which not only increases the stiffness of the arm part, but also can obtain larger end effector working space through a smaller drive stroke, and thus, the harvesting robot meets the requirements of large end effector working space when it clamps the camellia oleifera trunk. Robot can adjust the posture and position of the end effector by controlling the rotation of waist part and wrist part and the translation of arm part respectively. It makes the wrist part decrease by 2 degrees of freedom, lightens the burden of the arm, and also reduces the adverse effect of high-frequency vibration when the robot harvests camellia oleifera fruit. The screw theory and a kind of kinematics analysis method for hybrid robot were introduced. With the proposed method, firstly, the open chain from the waist part to the end effector of the manipulator was defined, which contained the vertical slider as major chain, and the kinematic analysis problem of hybrid robot was turned to the sub-problems of the kinematic analysis of an open major chain and a single closed chain; secondly, the kinematics equation of the closed chain and major chain was established, and then the conversion formula between the driving joint variable in closed chain and the passive joint variable in major chain was obtained by figuring out the equation of the closed chain; thirdly, the position of the end effector was got from the kinematics analysis of the closed chain by using the Lie group, Lie algebra, screw theory and product-of-exponential formula; and finally the positive kinematics equation for the position of the end effector was obtained from the synthesized results of the kinematic analysis of the major chain and closed chain. The first-order and second-order derivative of the position equation were the velocity equation and acceleration equation of the manipulator respectively. In order to verify the feasibility of the proposed method and the correctness of the kinematics equation, select a group of joint variable values, and then figure out the theoretical position coordinate of the end effector by putting the values into the kinematical equation in Matlab. A test platform for kinematics experiment was built, which consists of a three-dimensional position measuring instrument and a camellia oleifera fruit hybrid harvesting robot. The three-dimensional position measuring instrument was used to measure the actual position coordinates of the end effector directly, which was driven by the selected values. From the comparison between the theoretical and actual results, it was found that the maximum position error between kinematical equation resolution and actual position coordinate of the end effector was 10.4 mm, which was significantly smaller than 200 mm, the open size of the end effector. From the above experiment results, the correctness of the kinematics equation of the manipulator based on the proposed method was verified. Therefore, the application of the proposed method based on screw theory in kinematics analysis is beneficial for establishing the control method and trajectory planning.